Effects of Temperature and Light on Methane Production of Widespread Marine Phytoplankton

Methane (CH4) production in the ocean surface mixed layer is a widespread but still largely unexplained phenomenon. In this context marine algae have recently been described as a possible source of CH4 in surface waters. In the present study we investigated the effects of temperature and light intensity (including daylength) on CH4 formation from three widespread marine algal species Emiliania huxleyi, Phaeocystis globosa, and Chrysochromulina sp. Rates of E. huxleyi increased by 210% when temperature increased in a range from 10°C to 21.5°C, while a further increase in temperature (up to 23.8°C) showed reduction of CH4 production rates. Our results clearly showed that CH4 formation of E. huxleyi is controlled by light: When light intensity increased from 30 to 2,670 μmol m−2 s−1, CH4 emission rates increased continuously by almost 1 order of magnitude and was more than 1 order of magnitude higher when the daylength (light period) was extended from 6/18 hr light‐dark cycle to continuous light. Furthermore, light intensity is also an important factor controlling CH4 emissions of Chrysochromulina sp. and P. globosa and could therefore be a species‐independent regulator of phytoplankton CH4 production. Based on our results, we might conclude that extensive blooms of E. huxleyi could act as a main regional source of CH4 in surface water, since blooming of E. huxleyi is related to the seasonal increase in both light and temperature, which also stimulate CH4 production. Under typical global change scenarios, E. huxleyi will increase its CH4 production in the future.


Introduction
Huge amounts of methane (CH 4 ) are formed in the oceans, but only a small proportion is released to the atmosphere (Weber et al., 2019). In this context the biogeochemical cycle of CH 4 in the oceans is of great interest, and in particular, the frequently observed CH 4 production within the ocean surface mixed layer is challenging our previous understanding of biogeochemical CH 4 formation processes. Traditionally, it is thought that CH 4 in the oceans is either produced by geological processes (abiotic) or by methanogenic archaea (biotic). Because methanogenic archaea are strict anaerobic microorganism, their CH 4 production is limited to anoxic environments (Kirschke et al., 2013;Saunois et al., 2016;Thauer et al., 2008). However, there is growing evidence that CH 4 is also produced by organisms such as cyanobacteria (Bižić et al., 2020) and eukaryotes including plants (Keppler et al., 2006), fungi , lichens (Lenhart et al., 2015) and algae (Klintzsch et al., 2019;Lenhart et al., 2016), animals (Ghyczy et al., 2008), and humans  and even in the presence of oxygen.
The observation of CH 4 in freshwater and saline surface waters (often described as methane paradox) has recently received much attention although some studies already conducted four decades ago (Scranton, 1977;Scranton & Brewer, 1977;Scranton & Farrington, 1977) have reported about CH 4 supersaturation in the ocean mixed layer. Furthermore, many recent studies (Grossart et al., 2011;Günthel et al., 2019;Hartmann et al., 2020;Tang et al., 2016) have shown that CH 4 formation is not limited to saltwater but also occurs in freshwater lakes. Several hypotheses exist to explain CH 4 formation in oxygenated waters, and some of them will be discussed briefly. Methanogenic archaea living in anoxic environments of particles or fish and zooplankton guts might form CH 4 (de Angelis & Lee, 1994;Karl & Tilbrook, 1994;Schmale et al., 2018;Stawiarski et al., 2019;Zindler et al., 2013). The algal methabolit dimethylsulfoniopropionate (DMSP) and its degradation products dimethyl sulfide (DMS) or dimethyl sulfoxide (DMSO) could be precursors of both archaeal (Damm et al., 2008;Florez-Leiva et al., 2013) and bacterial produced CH 4 , when bacteria suffer under nitrogen deficiency (Damm et al., 2010). Moreover, photochemical degradation of DMS and acetone has been shown to produce CH 4 , but the reaction is limited to anoxic waters (Bange & Uher, 2005;Zhang, Xie, et al., 2015). In oligotrophic Pacific waters CH 4 formation might mainly related to the bacterial cleavage of methylphosphonates when supply of phosphorous is limited (del Valle & Karl, 2014;Karl et al., 2008;Metcalf et al., 2012;Repeta et al., 2016).
Phytoplankton might contribute to CH 4 production in both oxic marine and freshwater environments. The first indication of CH 4 production from phytoplankton was provided by culture experiments of the diatom species Thalassiosira pseudonana and the haptophyte species E. huxleyi (Scranton, 1977;Scranton & Brewer, 1977;Scranton & Farrington, 1977). Later on, many field studies have reported a relationship between CH 4 supersaturation and the occurrence of phytoplankton in lakes and oceans (e.g., Bogard et al., 2014;Conrad & Seiler, 1988;Damm et al., 2008;Grossart et al., 2011;Oudot et al., 2002;Owens et al., 1991;Rakowski et al., 2015;Tang et al., 2014;Weller et al., 2013;Zindler et al., 2013). Although a good statistical correlation was not observed in all previous studies (e.g., Brooks et al., 1981;Burke et al., 1983;Forster et al., 2009;Lamontagne et al., 1975;Watanabe et al., 1995), it was suggested that phytoplankton is one of the likely CH 4 sources. However, clear evidence of CH 4 formation from marine algae-examined in cultures of marine haptophytes-was only provided recently when Lenhart et al. (2016) and Klintzsch et al. (2019) applied stable isotope labeling experiments to unambiguously show that the three widespread marine algae such as E. huxleyi, Chrysochromulina sp., and Phaeocystis globosa indeed produce CH 4 per se and without the help of methanogenic archaea. Very recently, when using stable isotope labeling experiments and concentration measurements, it could be also shown that several freshwater algal species, including diatoms, cryptophytes, and green algae (Hartmann et al., 2020), but also several species of marine and limnic cyanobacteria (Bižić et al., 2020) emit CH 4 . Thus, both marine algae and cyanobacteria could significantly contribute to the commonly observed oceanic CH 4 supersaturation (Bižić et al., 2020;Klintzsch et al., 2019;Scranton, 1977). In summary, previous investigations mainly focused on explaining the sources for CH 4 in oxic surface waters; however, the effects of environmental parameters such as temperature, light intensity, or nutrient availability on CH 4 production from phytoplankton are still unknown.
In the present study we investigated the effects of temperature and light intensity (including daylength) on CH 4 formation from the three widespread marine algal species E. huxleyi, P. globosa, and Chrysochromulina sp. Emiliania huxleyi occurs in ocean worldwide except in the polar regions (McIntyre et al., 1970). The algal species develops large populations (blooms) in subpolar to temperate areas usually in summer time, especially under highly stratified conditions, when the mixed layer depth shallows due to increasing temperature. Blooming of E. huxleyi is then supported by high light intensity caused by shallow mixed layer depth and incidence light (Nanninga & Tyrrell, 1996;Raitsos et al., 2006;Tyrrell & Merico, 2004;Tyrrell & Taylor, 1996). Therefore, we have studied in detail CH 4 formation in relation to temperature (range from 10.1°C to 23.8°C), light intensity (30 to 2,670 μmol m −2 s −1 ), and daylength (period of light irradiation) during growth of E. huxleyi. We furthermore investigated the effect of light intensity on CH 4 formation by the two other widespread marine, but noncalcifying haptophytes P. globosa and Chrysochromulina sp. These two species can also form large blooms and are often found as main members in marine phytoplankton communities (Brown & Yoder, 1994;Schoemann et al., 2005;Thomsen, 1994). The results of E. huxleyi will be discussed with regard to their potential importance in marine environments during blooming. Finally, the observed CH 4 formation patterns of the three algal species will be evaluated on the basis of the CH 4 production potential (CH 4 -PP), which expresses differences in growth rates and thus the success of a species at the community level.

Experimental Setup
Emiliania huxleyi RCC1216 provided from the Roscoff Culture Collection (http://roscoff-culture-collection. org/, last access: 11 April 2020) were used to investigate the effect of temperature, light intensity, and daylength on CH 4 production rates. We performed an additional experiment to study the effect of light intensity on different algae species. Therefore, E. huxleyi and two other haptophytes P. globosa PLY 575 and Chrysochromulina sp. PLY 307 obtained from the Marine Biological Association of the United Kingdom (https://www.mba.ac.uk/facilities/culture-collection, last access: 11 April 2020) were studied. The cultures were maintained in quasi-exponential growth by frequent dilution with medium in order to keep them largely free of bacteria. All culture experiments were conducted under the use of sterile techniques. For a more detailed discussion about the potential interplay between algae and bacteria, we would like to refer the reader to the manuscript by Klintzsch et al. (2019). Briefly, Klintzsch et al. concluded that CH 4 production is clearly dependent on algal growth and that it is highly unlikely that bacteria alone are responsible for CH 4 production in the studied cultures. Each sample was taken at the end of the light period. Cultures were grown in batch mode (Langer et al., 2013). We used F/2 growth medium (Guillard & Ryther, 1962) that was based on sterile filtered (0.2 μm Ø pore size) North Sea seawater (sampled off Helgoland, Germany, 32 PSU). Cells were grown in crimped serum bottles (160 ml) filled with 140 ml medium and 20 ml headspace. Culture experiments were carried out with four independent repetitions. For determination of the CH 4 mixing ratio samples of 10 ml of headspace gas was sampled. The amount of produced CH 4 in culture group vials was calculated in respect to control groups. The culture and control group flasks were simultaneously sealed under ambient air and thus contained the same CH 4 background concentration. Please note that the produced CH 4 has been determined for the entire incubation flask-dissolved in the medium plus CH 4 of the headspace volume. For details on determination of CH 4 formation, please refer to section 2.2. For all experiments performed with algae and F/2 medium, the average CH 4 content in the cultures group at the end of incubation was higher than that found in the algae-free blanks as shown in Tables S1-S4. All growth rates and initial and final cell densities are given in the supporting information (Tables S1-S4).
Cultures were illuminated by cold white LED bulbs (LED Base Classic A100, Osram, Germany). The light spectrum of the LED bulbs is provided in Figure S1. The photosynthetic active radiation (PAR) was measured inside each incubation jar by using a light meter (ULM-500 Universal, WALZ, Germany) with a spherical quantum PAR sensor (US-SQS/L, WALZ, Germany). Temperature was logged by (UX120-006 M, HOBO, Germany).

Determination of CH 4 Mass
The CH 4 mass was determined at the end of the incubation period. In order to determine the CH 4 mass of the whole incubation flask (dissolved plus the CH 4 of the headspace volume), an aliquot (10 ml) of head space gas was taken from the incubation vials using a gastight syringe. In order to maintain headspace pressure when taking the headspace gas sample, an equivalent volume of seawater was injected into the flasks by syringe. The added volume was taken into account when determining the cell density (section 2.5). The sample gas was separated by gas chromatography using a GC-14B (Shimadzu, Japan) equipped with a 2 m column (Ø ¼ 3.175 mm inner diameter), packed with Molecular Sieve 5A 60/80 mesh from Supelco. Methane was recorded by a flame ionization detector (FID) and quantified (mixing ratio) by using two reference standards containing 9,837 and 2,192 parts per billion by volume (p.p.b.v) CH 4 . Mixing ratios were corrected for head space pressure. The latter was measured inside the incubation flask before gas sampling using a pressure meter (GMSD 1,3 BA, Greisinger). The CH 4 mass m CH4 ð Þ was determined by its mixing ratio (x CH4 ) and the ideal gas law (Equation 1), The dissolved CH 4 concentration was calculated by using the equation of Wiesenburg and Guinasso (1979).

Treatments of Alternating Temperature, Light Intensity, or Daylength at Cultures of E. huxleyi
To investigate the effect of temperature, light intensity, and daylength (daylength refers to the light period within a 24-hr light-dark cycle), three independent experiments with E. huxleyi were carried out in which one of the three parameters was varied as described in Figure 1. Within each experiment, the other two parameters were kept constant with 16/8-hr light-dark cycle;~500 μmol photons m −2 s −1 and 20°C, respectively. All treatments were carried out with four independent repetitions. Control groups contained F/2 medium only. Cultures were acclimated (≈10 generations) to the environmental conditions prior to the experiment. Cell density at inoculation varied between treatments depending on the growth rates under the given environmental conditions. Cultures of E. huxleyi were allowed to grow not more than 0.4 × 10 6 cells ml −1 (exponential phase) before they were harvested. The majority (>95%) of culture replicates reached final cell densities between 0.1 × 10 6 and 0.3 × 10 6 cells ml −1 (Tables S1-S3). Possible culture artifacts of CH 4 production rates, which could result from a cell density effect (Langer et al., 2013), were excluded for each investigated parameter by correlating the CH 4 production rates with the cell density on the harvest day ( Figure S2).

Treatments of Alternating, Light Intensity on Cultures of P. globosa, Chrysochromulina Sp., and E. huxleyi
The effect of light intensity was studied in two further haptophytes: P. globosa and Chrysochromulina sp. in addition to E. huxleyi. Two light intensities (427 ± 12 μmol m −2 s −1 and 1,165 ± 42 μmol m −2 s −1 ) with four replicates were applied, respectively. Cultures were grown under a 16/8 hr light-dark cycle and 20°C. Cultures were preadapted (≈10 generations) to light intensities before the experiment was started. The initial and final cell densities correspond to the exponential phase for each species (Klintzsch et al., 2019) and are given in the supporting information (Table S4). Control groups contained F/2 medium only.

Determination of Cell Density
For the determination of cell densities either a Fuschs-Rosenthal or Neubauer counting chamber (depending on cell density) was used. At least minimum of four aliquots of each culture sample were counted.

Determination of Growth and CH 4 Production Rates
All production rates were measured at exponentially growing cultures. For further information of measuring production rates from batch culture experiments, we refer to Klintzsch et al. (2019) and Langer et al. (2012Langer et al. ( , 2013. We calculated the growth rate (μ) from cell densities (N) of the beginning (t 0 , N 0 ) and end (t 1 , N 1 ) of the experiment (Equation 2).
The POC-based CH 4 production rates were calculated from the cellular organic carbon content (POC cell ). The latter was obtained from cell volume (V cell ) by using the carbon to volume relationship in Equation 3 according to Menden-Deuer and Lessard (2000): The cell volume was calculated from the cell diameter in light micrographs, which was measured by using the program ImageJ (Schindelin et al., 2012). We followed the recommendation of Olenina et al. (2006) and assumed a ball shape for calculating the cell volume for the three species investigated here.
The carbon-specific growth rate was calculated from the product of POC and growth rate μ (Equation 4): The CH 4 production rates were calculated by multiplying the growth rate μ with the corresponding cellular or POC-CH 4 quota, which was measured at the end of the experiment. The daily cellular CH 4 production rates (CH 4 P cell , ag CH 4 cell −1 day −1 , ag ¼ 10 −18 g) were calculated according to Equation 5: where m(CH 4 ) is the amount of CH 4 that was produced at the end of the experiment.
The daily cellular CH 4 production rates (CH 4 P POC , μg CH 4 g −1 POC day −1 ) were calculated from growth rate and CH 4 -POC quotas at the end of the experiment according to Equation 6.
The CH 4 production potential (CH 4 -PP) was calculated to scale variations in cellular production rates to community level. Detailed explanations for calculating the production potential (PP, which is not confined to CH 4 ) have been provided by  and . Please note that these authors have calculated the PP for CaCO 3 but the concept is the same for CH 4 . In accordance to the authors, the CH 4 -PP can be calculated for different growth periods, when a cellular standing stock for each time period is calculated from a given starting cell density (N 0 ). The related amount of produced CH 4 (CH 4 -PP) for each period of growth and respectively standing stock is the product of the cellular standing stock and CH 4 quota (Equation 7).
In the present study and in accordance to Klintzsch et al. (2019) the CH 4 -PP was calculated for a standing stock that is obtained after 7 days of growth starting with a single cell.
The sensitivity of growth, POC production and the rate of CH 4 formation to temperature were quantified by their activation energy (E a ), which is derived from the Arrhenius equation (Equation 8).

10.1029/2020JG005793
Journal of Geophysical Research: Biogeosciences where k ¼ reaction rate constant (here for growth, POC, or CH 4 production rate), A ¼ pre-exponential factor, R ¼ gas constant, and T ¼ temperature. The activation energies (E a ) of the rate can then be calculated by multiply the slope of the Arrhenius plot by −R, using a plot of ln (k) as function of T −1 .

Statistics
For each environmental factor (sections 3.1-3.3) the total data set of cellular and POC normalized CH 4 production was analyzed for statistical differences in the mean values among the treatment groups by using a one-way analysis of variance (ANOVA). Furthermore, within the individual culture experiments of E. huxleyi, P. globosa, and Chrysochromulina sp. (section 3.4), the mean values (cellular and POC normalized CH 4 production) of the two light intensities treatments (medium and high light) were compared by t tests.

Temperature Effect
Growth and POC production rates have more than doubled when increasing temperatures from 10.5°C to 21.5°C (Figures 2a and 2b). The optimum of growth and POC production was reached at 21.5°C (1.18 day −1 ; 13.7 ± 1.6 pg POC cell −1 day −1 ), while a further increase in temperature to 23.8°C led to a drastic reduction of about 50% for both growth and POC production rates. A similar pattern was observed for CH 4 production rates (cellular and POC normalized) as shown in Figures 2c and 2d. The POC normalized and cellular CH 4 production increased by 2.8-and 2.0-fold, respectively, when temperature increased by 11.4°C (from 10.1°C to 21.5°C). At 21.5°C the optimum of CH 4 production rates was reached (3.2 ± 0.6 μg CH 4 g −1 POC day −1 ; 37.4 ± 6.7 ag CH 4 cell −1 day −1 ). Further increase in temperature from 21.5°C to 23.8°C showed a reduction of CH 4 production by 40% and 35%, for POC normalized and cellular CH 4 production, respectively. Statistical analysis (ANOVA) confirmed the temperature dependence of CH 4 production with p values of 0.002 and <0.001 for POC normalized and cellular CH 4 production rates, respectively. After 1 week of growth the total amount of generated CH 4 is specified by the CH 4 -PP (Figure 2e). With increasing temperature (from 10.1°C to 21.5°C) the CH 4 -PP raised by 2 orders of magnitude (from 0.7 ± 0.2 to 124 ± 17.1 fg CH 4 ) before it declined drastically at 23.8°C (Figure 2e). Consequently, the optimum temperature (21.5°C) was identical for the five investigated parameters (Figure 2).

Light Intensity Effect
The growth and POC production rates of E. huxleyi increased drastically when light intensity increased from 30 to 171 μmol m −2 s −1 , and values remained relatively constant at higher light intensities in the range of 171-1,450 μmol m −2 s −1 (Figures 3a and 3b). However, increasing the light intensity to 2,670 μmol m −2 s −1 caused a clear reduction of both growth and POC production rates. From 30 to 171 μmol m −2 s −1 the growth rate increased fourfold (from 0.25 ± 0.09 to 1.00 ± 0.03 day −1 ) and POC production rates by over 1 order of magnitude (from 1.4 ± 0.48 to 17.1 ± 2.5 pg POC cell −1 day −1 ). The POC production was highest at 171 μmol m −2 s −1 and was similar at 872 μmol m −2 s −1 while rates were slightly smaller at 1452 μmol m −2 s −1 . The growth rate increased by 10% between 171 and 872 μmol m −2 s −1 but did not further change when reaching values of 1,450 μmol m −2 s −1 (1.13 ± 0.03 day −1 ). Thus, the optimum growth rate is reached at higher light intensities compared to the POC production rates. However, a further increase in light intensity up to 2,673 μmol m −2 s −1 led to a significant reduction (≈45%) of both POC production and growth rates. In contrast, the cellular and POC normalized CH 4 production rates increased steadily with increasing light intensity (Figures 3c and 3d). Methane formation (on a cell basis) was below the detection limit at 30 μmol m −2 s −1 but measurable (6.9 ± 9.5 ag CH 4 cell −1 day −1 ; 0.47 ± 0.63 pg POC cell −1 day −1 on average) between 60 and 171 μmol m −2 s −1 while at 872 μmol m −2 s −1 production rates strongly increased (23.9 ± 3.3 ag CH 4 cell −1 day −1 ; 1.6 ± 0.1 μg CH 4 g −1 POC day −1 ). From 872 to 2,670 μmol m −2 s −1 cellular and POC normalized CH 4 production increased by 4.2-and 5.1-fold up to 100 ± 12 ag CH 4 cell −1 day −1 and 6.8 ± 0.9 μg CH 4 g −1 POC day −1 , respectively. The light dependence of both cellular and POC-normalized CH 4 production was also indicated by statistical analysis (ANOVA; p < 0.001). The CH 4 -PP ( Figure 3e) increased with increasing light intensities by 2 orders of magnitude up to 152 ± 22 fg CH 4 at 1450 μmol m −2 s −1 and sharply decreased by on order of magnitude (to 9.9 ± 0.9 fg CH 4 ) at higher light intensity (2,670 μmol m −2 s −1 ). The optimum of CH 4 -PP is therefore in accordance with the optimum of growth rate.

Daylength Effect
The extension of the daylength (period of light irradiation) from 6 to 18 hr increased the growth rates 2.6-fold (from 0.32 ± 0.06 to 0.84 ± 0.04 day −1 ), while the growth rates remained constant when a period of continuous light (24 hr) was set ( Figure 4a). In contrast to the growth rate, an optimum of POC production rates was observed at 18 hr daylength and decreasing with longer irradiation period of 24 hr (Figure 4b). POC production rates increased between 6 and 18 hr daylength by 6.1-fold (from 1.6 ± 0.3 to 9.7 ± 0.9 pg POC cell −1 day −1 ) and declined by 22% at continuous light. Cellular and POC normalized CH 4 production rates increased from 6 hr daylength to continuous light period by 2 and 1 order of magnitude from 6.9 ± 3.4 to 186 ± 37 ag CH 4 cell −1 day −1 and 1.1 ± 0.6 to 21.0 ± 4.1 μg CH 4 g POC −1 day −1 , respectively (Figures 4c  and 4d). The dependence of CH 4 production on temperature was verified by statistical analysis (ANOVA), with cellular and POC normalized CH 4 production rates showing p values of p < 0.001 and p ¼ 0.004, respectively. The cellular and POC normalized CH 4 production was particularly enhanced by the 6 hr extension of the daylength between 18 hr and continuous light that accounted for 56% and 69% of the total increase in cellular and POC normalized CH 4 production, correspondingly. The CH 4 -PP increased constantly by over 2 orders of magnitude with longer light irradiation periods (from 0.19 ± 0.12 fg CH 4 at 6 hr light to 70.4 ± 23.1 fg CH 4 at continuous light, Figure 4e).

Comparison of Light Intensity Effects of E. huxleyi, Chrysochromulin Sp., and P. globosa
We compared growth and CH 4 formation patterns of the three algal species E. huxleyi, P. globosa, and Chrysochromulina sp. at moderate and high light intensities (429 μmol m −2 s −1 and 1,164 μmol m −2 s −1 ). The growth rates at both light intensities are shown in Figure 5a. At moderate light intensity the exponential growth rate μ was highest for E. huxleyi (0.90 ± 0.13 day −1 ) followed by P. globosa and Chrysochromulina sp. (with 0.57 ± 0.05 day −1 and 0.55 ± 0.04 day −1 , respectively). Growth rates of E. huxleyi and P. globosa remained constant at higher intensity, while growth rate of Chrysochromulina sp. declined by 29%. The POC production rates are shown in Figure 5b. At moderate light intensity the POC production rates of E. huxleyi and Chrysochromulina sp. were in the same range with 14.5 ± 0.5 and 15.5 ± 1.6 pg POC cell −1 day −1 , respectively, and were about three times higher than for P. globosa (4.1 ± 0.3 pg POC cell −1 day −1 ). The exposure to high light intensity led to a 31% lower POC production rate of Chrysochromulina sp. while rates of E. huxleyi and P. globosa remained constant. Thus, an increase in light intensity declined growth rate and POC production of Chrysochromulina sp., while that of E. huxleyi and P. globosa remained constant. Cellular CH 4 production rates of all investigated species were enhanced by increasing light intensities (Figure 5c). Cellular CH 4 production rates ranged from 17 ± 3.6 (Chrysochromulina sp.) to 27 ± 5.6 ag CH 4 cell −1 day −1 (E. huxleyi) at medium light. In response to higher light intensity the cellular CH 4 production rates increased by 2.6-fold (E. huxleyi) and about fivefold (P. globosa and Chrysochromulina sp.) resulting in a cellular CH 4 production rates ranged from 72.1 ± 3.3 (E. huxleyi) to 98.2 ± 30.8 ag CH 4 cell −1 day −1 (P. globosa). The response in cellular production rates to higher light intensity was also displayed by t test with p values of <0.001, 0.006, and 0.005 for E. huxleyi, P. globosa, and Chrysochromulina sp., respectively. The POC normalized CH 4 production rates increased with increasing light intensity (Figure 5d). Within the medium and high light intensities, the variation of POC normalized CH 4 production rates between

10.1029/2020JG005793
Journal of Geophysical Research: Biogeosciences species was greater than that of cellular CH 4 production rates. When CH 4 production rates were normalized to POC, rates of moderate light intensities were in a range of 0.58 ± 0.12 to 2.8 ± 1.6 μg CH 4 g −1 POC day −1 with Chrysochromulina sp. and P. globosa showing the lowest and highest rates, respectively. At high light intensity rates were about 2.8-fold (E. huxleyi) and 5.5-fold (P. globosa and Chrysochromulina sp.) greater than for those observed at moderate light intensity. These differences were also shown by t test with p values <0.001, 0.004, and <0.001 for E. huxleyi, P. globosa, and Chrysochromulina sp., respectively. The respectively rates ranged from 3.24 ± 0.78 (Chrysochromulina sp.) to 15.6 ± 2.2 μg CH 4 g −1 POC day −1 (P. globosa). All three species showed enhanced CH 4 -PP with the higher light intensity (Figure 5e). The increase from moderate to high light ranged between 2.4-fold (Chrysochromulina sp.) and 4.9-fold (P. globosa). However, the variation of the CH 4 -PP within the species is greater than that resulting from the different light treatments. The CH 4 -PP was 1 order of magnitude higher for E. huxleyi in comparison to the other two species (Figure 5e). This is in line with the higher growth rate of E. huxleyi.

Discussion
Previous studies indicated that several marine algae produce CH 4 (Klintzsch et al., 2019;Lenhart et al., 2016;Scranton, 1977;Scranton & Brewer, 1977;Scranton & Farrington, 1977), while the modulating influence of environmental parameters is unknown. Our results clearly show that CH 4 formation by E. huxleyi is influenced by temperature, light intensity, and the length of irradiation period. Furthermore, light intensity is also an important factor controlling emission rates of the two other marine algae Chrysochromulina sp. and P. globosa. We will first discuss the effects of environmental parameters on growth and POC normalized CH 4 production from a physiological perspective. Afterward, the effects of environmental parameters on laboratory CH 4 production rates of E. huxleyi are discussed in relation to their possible importance on populations (blooms) in marine environments. Finally, we discuss the impact of environmental parameters on CH 4 production in biogeochemical terms using the well-established but rarely applied concept of the PP (see Klintzsch et al., 2019, and references therein).

Temperature Effect on Growth and CH 4 Formation of E. huxleyi From a Physiological Perspective
Emiliania huxleyi occurs, except for the polar regions, in oceans worldwide and has the largest known temperature growth range (1-31°C) compared to other coccolithophores (McIntyre et al., 1970). The temperature response of growth rate is strain specific (Brand, 1982;Langer et al., 2009), and the optimum temperature for strain RCC1216 in this study tallies well with the published value (Langer et al., 2009). The growth curve (Figure 2a) exhibits the asymmetry typical for a temperature response. The ascending, shallow sloped, part of the curve is characterized by an accelerating effect of temperature on all biochemical reactions, whereas the descending, steep sloped, part is characterized by inactivation of enzymes, and denaturation of proteins and membranes (DeLong et al., 2017;Grimaud et al., 2017;Kingsolver, 2009). In accordance with this general concept of temperature effects on physiological processes, we observe a positive correlation of all analyzed physiological parameters with temperature up to 21.5°C (the optimum) followed by a negative correlation above this temperature. We conclude that CH 4 production is a normal physiological process as opposed to a heat stress response stemming from structural damage to cellular architecture. Please note that CH 4 production trends are identical, regardless of the normalization, that is, normalization to cell or POC (Figures 2c and 2d).
The ascending part of the temperature curve can be further analyzed using the Arrhenius equation. According to this equation (Equation 8), the thermal sensitivity of a chemical reaction is proportional to its activation energy. While the Arrhenius equation was originally used to describe chemical reactions, the equation might be also applied to describe the thermal sensitivity of biochemical reactions and biological growth rates, whereby high activation energies indicate high sensitivity to temperature (Gillooly et al., 2001;Grimaud et al., 2017). The calculated activation energies of growth rate, POC, and CH 4 production were 59, 41, and 63 kJ mol −1 , respectively. The growth rate and CH 4 production are therefore somewhat more sensitive to temperature than POC production is. The activation energy of CH 4 production is in the range of basic metabolic processes, indicating that CH 4 production in algae is not an abiotic process. For example, the average activation energy of respiration for a wide range of organisms, including microbes, plants, and animals, is between 40 and 71 kJ mol −1 (Gillooly et al., 2001). In addition, activation energies of most enzymatic reactions are in the range of 21 to 63 kJ mol −1 (Segel, 1993). By contrast abiotic CH 4 formation from thermal degradation experiments as described from dried soils usually showed higher activation energies above 70 kJ mol −1 (Jugold et al., 2012;Liu et al., 2019).

Light Intensity and Daylength Effects on Growth and CH 4 Formation of E. huxleyi From a Physiological Perspective
We grew E. huxleyi under a wide range of light intensities and daylengths.

Light Intensity
Our results demonstrate that CH 4 formation and growth rate of E. huxleyi is sensitive to light intensity. Under light-limited condition, growth rate increased sharply with increasing light intensity, leveled off at saturated light, and decreased at inhibiting light intensities (Figures 3a and 3b). This pattern of light intensity response (Figure 3a) is typical of phytoplankton cultures (Edwards et al., 2015, and reference inside). The optimum light intensity for POC production is lower than the one for growth rate, which was also observed by Trimborn et al. (2007). The growth rate of E. huxleyi was remarkably tolerant against high light intensities (≥1,500 μmol m −2 s −1 ), a phenomenon well documented in the literature (Balch et al., 1992;Harris et al., 2005;Loebl et al., 2010;Nanninga & Tyrrell, 1996;Nielsen, 1997;Trimborn et al., 2007). Interestingly CH 4 production was even more tolerant to high light, so much so that we could not determine the optimum light intensity. This is in notable contrast to the temperature response patterns described above. While we do not know the chain of events leading to this light intensity response, the response patterns suggest that, first, CH 4 production is a light dependent process and, second, photo-inhibition of growth rate and POC production do not impair CH 4 production. The latter is particularly intriguing because it seems to suggest a decoupling of CH 4 production from photosynthetic production of both energy equivalents and putative CH 4 precursors originating in the POC pool. This warrants further, more detailed, physiological studies into the nature of the light dependency of CH 4 production.

Daylength
The CH 4 formation and growth rate were furthermore controlled by daylength. The growth rate showed a saturation curve (Figure 4a). This pattern is similar to results on E. huxleyi reported by Paasche (1967) and even other phytoplankton species (e.g., Bouterfas et al., 2006). By contrast, growth rates of E. huxleyi have been reported to be independent of daylength (Nielsen, 1997) or to be inhibited by continuous light (Van Rijssel & Gieskes, 2002). The response to daylength has been suggested to be strain specific (Bretherton et al., 2019) and is dependent on other environmental parameters, for example, on seawater CO 2 concentration (Bretherton et al., 2019;Zhang, Bach, et al., 2015) and light quality (Glover et al., 1987). This could be one reason why the response of growth in relation to daylength differs between studies. In our study POC production decreased at continuous light while growth rate did not (Figure 4b). While growth rate might be inhibited by a lack of a dark period rather than by photoinhibition (Brand & Guillard, 1981), the decline of POC production at continuous light could partly be due to photoinhibition. Please note (see also above) that the response pattern of CH 4 production is independent of the normalization (cell or POC). Interestingly, it is again CH 4 production that neither levels off nor shows inhibition at continuous light. This observation reinforces the pattern described above, namely, POC production declines while CH 4 production increases further. This "double dependency" of CH 4 production on light, that is, both light intensity and daylength, renders light-dependent processes the prime target for further elucidating the mechanism of CH 4 production in E. huxleyi.

Light Intensity Effects on Growth and CH 4 Formation of E. huxleyi, Chrysochromulin Sp., and P. globosa
Methane production was light dependent in P. globosa and Chrysochromulina sp. too. Emiliania huxleyi and P. globosa showed a greater light tolerance with respect to growth rate and POC production than Chrysochromulin sp. (Figures 5a and 5b). In contrast to POC production and growth rate of Chrysochromulina sp., CH 4 production was not inhibited and once again confirms the remarkable light dependency of CH 4 production. The increase in CH 4 production with light intensity in P. globosa and Chrysochromulina sp. was even higher than that of E. huxleyi. However, in each case there was a positive correlation of CH 4 production and light intensity, which could therefore be a common feature of different phytoplankton taxa. This hypothesis is supported by recent findings of Bižić et al. (2020), who investigated different cyanobacterial species that are found in phytoplankton communities of ocean and lakes. While cyanobacteria produce CH 4 in light and dark phase, the CH 4 production rates elevated during the light phase.

Potential Relevance for CH 4 Production of E. huxleyi Populations in the Field
Large blooms of E. huxleyi typically occur at subpolar to temperate areas in the summer months, when the water is highly stratified, due to the seasonal increase in temperature and light intensity (Iglesias-Rodríguez et al., 2002;Nanninga & Tyrrell, 1996;Raitsos et al., 2006;Tyrrell & Merico, 2004). The occurrence of E. huxleyi in field is therefore correlated with high solar radiation, shallow mixed layer depth, and increased sea surface temperature (SST; Raitsos et al., 2006). We compare the reported environmental conditions that support E. huxleyi growth in the field with those that stimulate CH 4 production in our laboratory grown cultures to assess whether CH 4 formation by E. huxleyi could be of ecological relevance.
Emiliania huxleyi grows at water temperatures between 1°C and 31°C in field and has the largest temperature growth range of all coccolithophores (McIntyre et al., 1970). The wide temperature range results from the adaptation of individual strains to narrower temperature ranges in cold or warm water masses (Brand, 1982;Langer et al., 2009). The temperature range of the incubation experiments includes the range of the seasonal variation in the SST of the Tasman Sea off New Zealand where the investigated strain (RCC1216) has been isolated and is therefore of ecological relevance. The monthly mean SST of the Tasman Sea off New Zealand ranges from 13.5°C to 18.7°C between coldest and warmest month (time period 2007-2017, https://statisticsnz.shinyapps.io/sea_surface_temperature_oct19/, last access: 11 April 2020). With a temperature increase from 10.1°C to 21.5°C cellular CH 4 production rates would double according to laboratory derived rates (section 3.1). Inhibition of growth rate and CH 4 production due to heat stress is less likely in the field, since the mean SST of the warmest month is below the optimum temperature of the investigated strain. This is in line with literature data showing that most strains grow below their optimal growth condition in field (Langer et al., 2009;Rosas-Navarro et al., 2016).
Our laboratory data also suggest that longer light irradiation periods during summer could have a stimulating effect on CH 4 production, especially on E. huxleyi populations in subpolar regions, where daylength changes dramatically between winter and summer. For example, the CH 4 production would increase by a factor of 5 due to a daylength increase from 6 hr in winter to 18 hr in summer (section 3.3). Emiliania huxleyi usually blooms in the North Atlantic in June and July at high light intensity in the surface layer, which is caused by strong sunlight and shallow mixed layer depth (10-20 m; Nanninga & Tyrrell, 1996;Raitsos et al., 2006;Tyrrell & Merico, 2004;Tyrrell & Taylor, 1996). For instance, light intensities of 935 and 1,140 μmol m −2 s −1 were measured in E. huxleyi blooms in the field. In addition, long-term observations in a mesocosm in a Norwegian fjord have shown that E. huxleyi blooms at light intensities between >530 and 1,200 μmol m −2 s −1 , whereas the mean light intensities in the surface layer are~63% and 43% of the incident light intensity at 10 and 20 m mixed layer depth, respectively (Nanninga & Tyrrell, 1996, and reference inside). Thus, the E. huxleyi cells would have been exposed to light intensities between 228 and 756 μmol m −2 s −1 in the mixed layer, which falls within the light intensity range of our incubation experiments (section 3.2). Since the CH 4 production of E. huxleyi increased linearly with the light intensities in culture experiments, the high light intensities in the surface layer could also support the CH 4 formation of E. huxleyi in the field. Judging from our laboratory CH 4 production and the reported light intensity range where blooms typically occur, the CH 4 production could vary by a factor of 4. It can be concluded that light intensity will considerably affect sea surface water CH 4 production of E. huxleyi in field.

The Biogeochemical Perspective: Methane PP (CH 4 -PP)
While the considerations in section 4.4 apply to the behavior of field populations on the cellular level, they are not appropriate for assessing the biogeochemical significance of this behavior. Several recent studies have emphasized that the PP (see section 2 for calculation), as opposed to the cellular production, is the relevant parameter for biogeochemical assessments Klintzsch et al., 2019;Kottmeier et al., 2016;Marra, 2002;Schlüter et al., 2014). We calculated the CH 4 -PP of E. huxleyi for different temperature, light intensity, and daylength conditions. For all three parameters, the CH 4 -PP increases toward the optimum, as does the cellular CH 4 production, but the increase in CH 4 -PP was by 1 order of magnitude higher than for the cellular CH 4 production (Figures 2c, 3c, and 4c and Figures 2e, 3e, and 4e). This illustrates the importance of using the PP when considering the biogeochemical impact of changing environmental conditions. Another such illustration is the strong contrast between the light intensity response patterns of cellular, or POC normalized, CH 4 production, and the CH 4 -PP. The sharp decline in CH 4 -PP at the highest light intensity is not reflected in the cellular CH 4 production curve. However, in the field this difference is of minor importance because the highest light intensity used here, 2,700 μmol m −2 s −1 , is considerably higher than even peak light intensities observed in typical E. huxleyi blooms (~1,200 μmol m −2 s −1 ; see references above). It is noteworthy that the decline in CH 4 -PP at the highest temperature tested here is also of little relevance in the field because E. huxleyi usually grows at suboptimal temperatures in the field, a situation that will also not change in the foreseeable future, despite global warming (Rosas-Navarro et al., 2016). It is concluded first that the CH 4 -PP of E. huxleyi in the field will be maximal in midsummer when E. huxleyi typically blooms. Second, global change will increase the CH 4 -PP of E. huxleyi through both warming and increased stratification entailing higher light intensities in the surface layer. Compared to the other two tested haptophytes, E. huxleyi has the highest CH 4 -PP, a difference not mirrored in cellular CH 4 production: This is yet another example of the importance of using the PP when considering the biogeochemical impact of CH 4 formation by phytoplankton. As a general caveat it should be noted that the above conclusions are confined to our experimental conditions. Conditions in the field will include other factors such as grazing. This inevitable limitation of experimental data is the price one has to pay for discovering relationships between environmental parameters and the performance of an organism. However, this does not mean that our data are never directly applicable to the field situation as illustrated by the good match of satellite data and E. huxleyi calcite production potential reported by .

Conclusions
We have determined the CH 4 production of three haptophytes under varying environmental conditions and conclude the following: 1. Temperature, light intensity, and daylength influence CH 4 production. 2. CH 4 production is strongly light dependent; even increasing with light intensity when growth rate and POC production are photoinhibited. 3. The biogeochemically relevant parameter CH 4 -PP increased with temperature, light intensity, and daylength over the range typical for present-day seasonality and global change predictions for the coming century. 4. E. huxleyi has a considerably higher CH 4 -PP than P. globosa and Chrysochromulina sp.