Methane production by three widespread marine phytoplankton species: release rates, precursor compounds, and potential relevance for the environment

Methane (CH₄) production within the oceanic mixed layer is a widespread phenomenon, but the underlying mechanisms are still under debate. Marine algae might contribute to the observed CH₄ oversaturation in oxic waters, but so far direct evidence for CH₄ production by marine algae has only been provided for the coccolithophore Emiliania huxleyi.

In the present study we investigated, next to E. huxleyi, other widespread haptophytes, i.e., Phaeocystis globosa and Chrysochromulina sp. We performed CH₄ production and stable carbon isotope measurements and provide unambiguous evidence that all three investigated marine algae are involved in the production of CH₄ under oxic conditions. Rates ranged from 1.9±0.6 to 3.1±0.4 µg of CH₄ per gram of POC (particulate organic carbon) per day, with Chrysochromulina sp. and E. huxleyi showing the lowest and highest rates, respectively. Cellular CH₄ production rates ranged from 16.8±6.5 (P. globosa) to 62.3±6.4 ag CH₄ cell⁻¹ d⁻¹ (E. huxleyi; ag = 10⁻¹⁸ g). In cultures that were treated with ¹³C-labeled hydrogen carbonate, δ¹³CH₄ values increased with incubation time, resulting from the conversion of ¹³C–hydrogen carbonate to ¹³CH₄. The addition of ¹³C-labeled dimethyl sulfide, dimethyl sulfoxide, and methionine sulfoxide – known algal metabolites that are ubiquitous in marine surface layers – resulted in the occurrence of ¹³C-enriched CH₄ in cultures of E. huxleyi, clearly indicating that methylated sulfur compounds are also precursors of CH₄. By comparing the algal CH₄ production rates from our laboratory experiments with results previously reported in two field studies of the Pacific Ocean and the Baltic Sea, we might conclude that algae-mediated CH₄ release is contributing to CH₄ oversaturation in oxic waters. Therefore, we propose that haptophyte mediated CH₄ production could be a common and important process in marine surface waters.

1 Introduction

Methane (CH₄), the second most important anthropogenic greenhouse gas after CO₂, is the most abundant reduced organic compound in the atmosphere and plays a central role in atmospheric chemistry (Denman et al., 2007; Kirschke et al., 2013; Lelieveld et al., 1998). The mixing ratio of CH₄ in the atmosphere has been increasing dramatically from preindustrial values of about 715 parts per billion by volume (ppbv) to about 1868 ppbv (October 2018, NOAA). The global atmospheric CH₄ budget is determined by the total emission (540–568 Tg CH₄ yr⁻¹) of various sources from terrestrial and aquatic surface areas that are balanced primarily by one major sink (hydroxyl radicals) in the atmosphere. The world's oceans are considered to be a minor source of CH₄ to the atmosphere (1 %–3 %, Saunois et al., 2016). However, in recent years the widespread occurrence of in situ CH₄ production in the ocean mixed layer has received much attention, since CH₄ formation in the oxygenated ocean mixed layer challenges the paradigm that biological methanogenesis is a strictly anaerobic process.

Methane is primarily formed by the degradation of buried organic matter under heat and pressure (thermogenic) inside the Earth's crust or produced by the incomplete combustion of biomass (pyrogenic). However, CH₄ resulting from microbial processes, carried out by methanogenic archaea under anoxic conditions in soils and sediments or the digestion system of ruminants, is categorized as biogenic or microbial (Kirschke et al., 2013). In contrast to these well-known sources, recent studies have confirmed direct CH₄ release from eukaryotes, including plants, animals, fungi, lichens, and the marine alga Emiliania huxleyi even in the absence of methanogenic archaea and in the presence of oxygen or other oxidants (Keppler et al., 2006; Ghyczy et al., 2008; Lenhart et al., 2012, 2015b, 2016). A very recent study also confirmed Cyanobacteria as CH₄ producers, suggesting that CH₄ production occurs in all three domains of life (Bizic-Ionescu et al., 2018a). These novel sources, from the domains Eucarya and Bacteria, might be classified as biotic non-archaeal CH₄ (Boros and Keppler, 2018).

In situ CH₄ production in oxygenated surface waters in the marine environment was first reported by Scranton and Farrington (1977) and Scranton and Brewer (1977) and some decades later also for lakes (Grossart et al., 2011). Significant quantities of CH₄, produced in upper oxic waters near the air–water interface, might overcome oxidation and thus significantly contribute to CH₄ fluxes from aquatic environments to the atmosphere (Bogard et al., 2014). It turned out that in situ CH₄ production in the upper oxic waters is a common feature of both oceans and lakes (Forster et al., 2009; Reeburgh, 2007; Tang et al., 2014; Donis et al., 2017; Bižić-Ionescu et al., 2018b; Bange et al., 1994). These results have stimulated the scientific community to study in more detail the phenomenon of CH₄ occurrence in oxygenated surface waters. In this context, emissions from Cyanobacteria or algae might help to explain the phenomenon of dissolved CH₄ oversaturation. In addition, it has been suggested that CH₄ might be produced under phosphorus limitation by the bacterial cleavage of methylphosphonate (MPn) in oligotrophic Pacific regions (Karl et al., 2008; Metcalf et al., 2012; Repeta et al., 2016; Valle and Karl, 2014).

In contrast to this apparently non-oxygen-sensitive pathway, many other studies have identified the “traditionally” archaeal methanogenesis in anoxic microenvironments as a CH₄ source. Floating particles (Karl and Tilbrook, 1994), as well as the digestive tracts of zooplankton (de Angelis and Lee, 1994; Stawiarski et al., 2019; Schmale et al., 2018) or fishes (Oremland, 1979), have been found as anoxic micro-niches for methanogens. It has been suggested that some methanogens might be active under oxic conditions by being equipped with enzymes to counteract the effects of molecular oxygen during methanogenesis (Angel et al., 2011). Potential substrates for methylotrophic methanogens in such micro-niches are the algae metabolites dimethylsulfoniopropionate (DMSP) and their degradation products dimethyl sulfide (DMS) or dimethyl sulfoxide (DMSO) (Zindler et al., 2013; Damm et al., 2008; Florez-Leiva et al., 2013). Furthermore, DMSP might also be converted to CH₄ by nitrogen-limited bacteria (Damm et al., 2010, 2015). However, in coastal waters where DMS and DMSP production is enhanced, CH₄ was found to be mainly related to sedimentary sources (Borges et al., 2018).

In contrast to microbial processes, which are considered to be driven by enzymes, CH₄ might also be derived by the chemical reaction of chromophoric dissolved organic matter (CDOM) and DMS induced by UV or visible light under both oxic and anoxic conditions (Zhang et al., 2015). A similar photochemical CH₄ formation was described earlier for acetone by Bange and Uher (2005), but the production of CH₄ from acetone was considered negligible under oxic conditions.

Under highly oxidative conditions nonheme iron-oxo (IV) species catalyze CH₄ formation from methyl thioethers and their sulfoxides (Althoff et al., 2014; Benzing et al., 2017). Iron-oxo species are intermediates in a number of biological enzymatic systems (Hohenberger et al., 2012). Thus, marine algae containing elevated concentrations of methyl thiolethers and their sulfoxides such as DMSP, DMSO, methionine (MET), or methionine sulfoxide (MSO) might be biochemical reactors for non-archaeal CH₄ production as was already proposed by Lenhart et al. (2016) and Keppler et al. (2009). Marine phytoplankton plays a central role in the global carbon cycle: approximately half of Earth's primary production is carried out by marine phytoplankton (Field et al., 1998). In this context it is important to mention that almost 40 years ago researchers (Scranton and Brewer, 1977; Scranton and Farrington, 1977; Scranton, 1977) already mentioned the possibility of the in situ formation of CH₄ by marine algae, since CH₄ production was examined in cultures of E. huxleyi and Thalassiosira pseudonana. Furthermore, direct isotopic evidence for CH₄ production by marine algae in the absence of methanogenic archaea has only been provided for E. huxleyi (Lenhart et al., 2016). Based on the application of stable carbon isotope techniques, it could be clearly shown that both hydrogen carbonate and a position-specific ¹³C-labeled MET were carbon precursors of the observed CH₄ production. However, it remains unclear whether CH₄ production also occurs among other marine algae and if there are also other carbon precursors involved in the formation process.

In the present study we investigated, next to the coccolithophore E. huxleyi, two other marine, non-calcifying haptophytes, namely Phaeocystis globosa and Chrysochromulina sp., for CH₄ formation. The investigated species are all bloom-forming and often found as dominant members in marine phytoplankton communities worldwide (Schoemann et al., 2005; Thomsen, 1994; Brown and Yoder, 1994). Furthermore, they are well-known for their high DMSP, DMS, and DMSO productivity (Liss et al., 1994; Keller, 1989; Holligan et al., 1993; Stefels et al., 2007; Matrai and Keller, 1993). We therefore conducted stable isotope experiments using ¹³C-labeled DMS, MSO, and DMSO to identify potential methyl group precursor compounds that eventually lead to CH₄ production. Finally, we discuss the laboratory CH₄ production rates in relation to their potential significance in marine environments.

2 Material and methods

2.1 Cultures and culture conditions

Three algal species, E. huxleyi RCC1216 obtained from the Roscoff Culture Collection (http://roscoff-culture-collection.org/, last access: 21 October 2019), 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: 21 October 2019), were studied. In order to keep non-axenic algae cultures largely free of bacteria, the cultures were diluted regularly, resulting in quasi-constant exponential algal growth while minimizing bacterial cell density. All incubation experiments were carried out in controlled and sterile laboratory conditions under a 16 ∕ 8 h light–dark cycle at a light intensity of 350 µmol photons m⁻² s⁻¹ and a temperature of 20 ^∘C. All samples were taken at the end of the light cycle. Monoclonal cultures were grown in full-batch mode (Langer et al., 2013) in sterile filtered (0.2 µm ∅ pore size) natural North Sea seawater (sampled off Helgoland, Germany) enriched in nutrients according to F/2 medium (Guillard and Ryther, 1962). The initial dissolved inorganic carbon (DIC) of the F/2 medium was 2152±6 µmol L⁻¹ (measured by Shimadzu TOC-V CPH). The DIC value falls within the range of typical DIC concentrations of North Sea seawater.

2.2 Determination of cell densities

Cell densities were determined from four aliquots of each culture sample using either a Fuschs–Rosenthal or Neubauer counting chamber, depending on cell density.

2.3 Incubation with ¹³C-labeled hydrogen carbonate

To investigate CH₄ production by algal cultures, borosilicate glass bottles (Schott, Germany) filled with 2.0 L of 0.2 µm filtered F/2 medium and with 0.35 L headspace volume were used in our investigations of Chrysochromulina sp. and P. globosa. For the investigations of E. huxleyi 0.85 L medium and 0.4 L headspace volume were used (Schott, Germany). The vials were sealed airtight with lids (GL 45, PP, two-port, Duran Group) equipped with one three-way port for liquid and a second port fitted with a septum for gas sampling. For measurements of the mixing ratio and stable carbon isotope value of CH₄ (δ¹³C−CH₄) samples of headspace (20 mL) were taken from each vial. Afterwards, samples (2 mL) for determining cell densities were taken. In order to maintain atmospheric pressure within the vial, the surrounding air was allowed to enter via the three-way port and through a sterile filter to avoid biological contamination. The inflow of surrounding air was taken into consideration when CH₄ production was calculated. Cultures that were studied during the incubation were inoculated from a pre-culture grown in dilute-batch mode (Langer et al., 2009). To investigate algal-derived CH₄ formation six vials were inoculated with algae and another six vials contained medium only. In addition, three vials of each group were treated with ¹³C–hydrogen carbonate (${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$) to investigate CH₄ formation by measuring stable carbon isotope values of CH₄. Four different treatments were used: medium either with ${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$ (medium + ${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$) or without (medium, data not available) and cultures supplemented either with ${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$ (medium + culture + ${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$) or without (medium + culture). The different treatments and the number of replicates for the experiments with Chrysochromulina sp. and P. globosa are provided in Fig. 1. Please note that stable isotope measurements using ${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$ were not performed for E. huxleyi as evidence for the isotope labeling of CH₄ was already provided by Lenhart et al. (2016). To study the CH₄ formation of E. huxleyi by measuring headspace concentration, three replicates (culture and medium group, n=3) were used.

Figure 1 Experimental setup for measuring CH₄ formation by Chrysochromulina sp. and P. globosa. Methane formation was investigated by concentration measurements within six vials containing either algae or medium only (left column). For stable isotope measurements of CH₄, ¹³C-labeled hydrogen carbonate (${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$) was added to three vials of both groups (right column).

The overall incubation time was 9, 11, and 6 d for Chrysochromulina sp., P. globosa, and E. huxleyi, respectively. Headspace and liquid samples were collected on a daily basis for E. huxleyi and in 2–3 d intervals from cultures of Chrysochromulina sp. and P. globosa. The incubation time and sampling intervals varied between species because of variations in the growth rate and the cell density in the stationary phase. Cell densities were plotted versus time, and the exponential growth rate (μ) was calculated from exponential regression using the natural logarithm (Langer et al., 2013). The exponential growth phase (from which μ was calculated) was defined by the cell densities which corresponded to the best fit (r²>0.99) of the exponential regression. This was done by using the first three (Chrysochromulina sp. and E. huxleyi) or four data points (P. globosa) of the growth curve. For stable carbon isotope experiments 48.7 µmol L⁻¹ NaH¹³CO₃ in final concentration was added to the F/2 medium. The added amount of NaH¹³CO₃ corresponds to 2 % of the DIC of the North Sea seawater (2152±6 µmol L⁻¹), resulting in a theoretically calculated δ¹³C value of DIC of $+\mathrm{2014}\pm \mathrm{331}$ ‰ . To determine the δ¹³C−CH₄ values of the source, the Keeling plot method was applied (Keeling, 1958). For a detailed discussion of the Keeling plot method for determination of the isotope ratio of CH₄ in environmental applications, please refer to Keppler et al. (2016). Oxygen concentration was monitored daily (using inline oxygen sensor probes; PreSens, Regensburg) at the end of the light cycle (Fig. S1 in the Supplement).

2.4 Determination of CH₄ production rates

Since the experiment in the Sect. 2.3 was not designed to obtain POC quotas (POC: particulate organic carbon), we conducted an additional experiment. To best compare the CH₄ formation rates of the three algae species it is necessary to obtain exponential growth to ensure constant growth rates and constant (at a given time of day) cellular POC quotas over the course of the experiment. Exponential growth is a prerequisite for calculating production on the basis of growth rate and quota (here CH₄ quota). The point is a general, technical one and is not confined to CH₄ production. The studies by Langer et al. (2012, 2013) discuss this point in the context of batch culture experiments. Briefly, production on this account is the product of growth rate and quota (e.g., CH₄, calcite, organic carbon). Production here is an integrated value, typically over many cell divisions. For this calculation of production to be meaningful a constant growth rate is required. The exponential growth phase fulfills this criterion, whereas the transition phase and the stationary phase do not. Therefore, production cannot be calculated meaningfully in the non-exponential phases. The problem can, however, be minimized by using small increments (1 d) because growth rate can be regarded as quasi-constant (see also Lenhart et al., 2016). The CH₄ production rates can be calculated by multiplying the growth rate μ with the corresponding cellular or POC–CH₄ quota that was measured at the end of the experiment. For this additional experiment the cultures were grown in 160 mL crimped serum bottles filled with 140 mL medium and 20 mL headspace (n=4). Oxygen concentration was monitored (using inline oxygen sensor probes; PreSens, Regensburg) at the end of each light and dark cycle (Fig. S2). The growth rate (μ) was calculated from cell densities of the beginning and end of the experiment according to Eq. (1):

$$\begin{array}{}\text{(1)}& \mathit{\mu }={\displaystyle \frac{\mathrm{Ln}\left(N_{\mathrm{1}}\right)-\mathrm{Ln}\left(N_{\mathrm{0}}\right)}{\left(t_{\mathrm{1}}-t_{\mathrm{0}}\right)}},\end{array}$$

where N₀ and N₁ are the cell densities at the beginning (t₀) and end of the experiment (t₁). The daily cellular CH₄ production rates (CH₄P_cell, ag CH₄ cell⁻¹ d⁻¹, ag = 10⁻¹⁸ g) were calculated according to Eq. (2):

$$\begin{array}{}\text{(2)}& {{\mathrm{CH}}_{\mathrm{4}}\mathrm{P}}_{\mathrm{cell}}=\mathit{\mu }\times {\displaystyle \frac{m\left({\mathrm{CH}}_{\mathrm{4}}\right)}{\mathrm{cell}}},\end{array}$$

where m(CH₄) is the amount of CH₄ that was produced at the end of the experiment.

To calculate POC-based CH₄ production rates the cellular organic carbon content (POC_cell) was derived from cell volume (V_cell) by using the Eq. (3) according to Menden-Deuer and Lessard (2000):

$$\begin{array}{}\text{(3)}& {\mathrm{POC}}_{\mathrm{cell}}=\mathrm{0.216}\times V_{\mathrm{cell}}^{\mathrm{0.939}}.\end{array}$$

The cell volume was determined by measuring the cell diameter in light micrographs using the program ImageJ (Schindelin et al., 2012). According to Olenina (2006) a ball shape can be assumed for calculating the cell volume for the three species investigated here. The daily cellular CH₄ production rates (CH₄P_POC, µg CH₄ g⁻¹ POC d⁻¹) were calculated from growth rate and CH₄–POC quotas at the end of the experiment according to Eq. (4).

$$\begin{array}{}\text{(4)}& {{\mathrm{CH}}_{\mathrm{4}}\mathrm{P}}_{\mathrm{POC}}=\mathit{\mu }\times {\displaystyle \frac{m\left({\mathrm{CH}}_{\mathrm{4}}\right)}{\mathrm{POC}}}\end{array}$$

The CH₄ production potential (CH₄–PP) was used to translate differences in cellular production rates to the community level. According to Gafar et al. (2018), the CH₄–PP can be calculated for different periods of growth by calculating a cellular standing stock for each time period from a known starting cell density (N₀) (whereby constant exponential growth is assumed). The corresponding amount of produced CH₄ (CH₄PP) for each period of growth and standing stock is the product of the cellular standing stock and CH₄ quota (Eq. 5).

$$\begin{array}{}\text{(5)}& {\mathrm{CH}}_{\mathrm{4}}\mathrm{PP}=N_{\mathrm{0}}\times e^{\mathit{\mu }\times t}\times {\displaystyle \frac{m\left({\mathrm{CH}}_{\mathrm{4}}\right)}{\mathrm{cell}}}\end{array}$$

In the present study the CH₄–PP was calculated for a standing stock that is obtained after 7 d of growth starting with a single cell.

2.5 Incubation with ¹³C labeled DMS, DMSO, and MSO

The sulfur-bonded methyl group(s) in DMS, DMSO, and MSO were investigated as precursors for algal-derived CH₄ in an incubation experiment with E. huxleyi. For all tested compounds only the C atom of the sulfur-bonded methyl group(s) was labeled with ¹³C (R–S–¹³CH₃, 99 %). A final concentration of 10 µM was used for each compound. The different treatments to investigate potential CH₄ formation by ¹³C₂–DMS, ¹³C₂–DMSO, ¹³C–MSO are provided in Fig. 2. Three independent replicates and repeated measurements over time were used. Headspace and vial size were analogous to the experiment described in Sect. 2.3 for E. huxleyi. Samples were taken daily during an overall incubation time of 6 d.

Figure 2 Experimental setup to investigate potential precursor compounds of CH₄. Dimethyl sulfide (¹³C₂–DMS), dimethyl sulfoxide (¹³C₂–DMSO), and methionine sulfoxide (¹³C–MSO) were added to the vials containing either a culture of E. huxleyi or medium only. For all tested compounds only the carbon atom of the sulfur-bonded methyl group(s) was labeled with ¹³C.

2.6 Determination of CH₄ mass

A 5 mL gas sample was collected from the headspace of the vials using a gastight Hamilton gas syringe. The sample was analyzed by a gas chromatographer (GC-14B, Shimadzu, Japan; column: 2 m, ∅=3.175 mm inner diameter, high-grade steel tube packed with molecular sieve 5A 60/80 mesh from Supelco) equipped with a flame ionization detector (FID). Quantification of CH₄ was carried out by comparison of the integrals of the peaks eluting at the same retention time as that of the CH₄ authentic standard using two reference standards containing 9837 and 2192 ppbv. Mixing ratios were corrected for headspace pressure, which was monitored using a pressure-measuring device (GMSD 1, 3 BA, Greisinger). The CH₄ mass ($m_{{\mathrm{CH}}_{\mathrm{4}}}$) was determined by its mixing ratio ($x_{{\mathrm{CH}}_{\mathrm{4}}}$) and the ideal gas law (Eq. 6):

$$\begin{array}{}\text{(6)}& m_{{\mathrm{CH}}_{\mathrm{4}}}=M_{{\mathrm{CH}}_{\mathrm{4}}}\times x_{{\mathrm{CH}}_{\mathrm{4}}}{\displaystyle \frac{p\times V}{R\times T}},\end{array}$$

where $M_{{\mathrm{CH}}_{\mathrm{4}}}$ is molar mass, p is pressure, T is temperature, R is the ideal gas constant, and V is volume.

The dissolved CH₄ concentration was calculated by using the equation of Wiesenburg and Guinasso (1979).

2.7 GC-C-IRMS measurements

Stable carbon isotope values of the CH₄ of headspace samples were analyzed by gas chromatography–stable isotope ratio mass spectrometry (GC-C-IRMS; Deltaplus XL, Thermo Finnigan, Bremen, Germany). All δ¹³C–CH₄ values were corrected using two CH₄ working standards (Isometric Instruments, Victoria, Canada) with values of $-\mathrm{23.9}\pm \mathrm{0.2}$ ‰ and $-\mathrm{54.5}\pm \mathrm{0.2}$ ‰. The results were normalized by two-scale anchor calibration according to Paul et al. (2007). The average standard deviation of the analytical measurements was in the range of 0.1 ‰ to 0.3 ‰ (based on three repeated measurements of CH₄ working standards). All δ¹³C–CH₄ values are expressed in the conventional δ notation, in per mille (‰) vs. Vienna Pee Dee Belemnite (VPDB), using Eq. (7).

$$\begin{array}{}\text{(7)}& {\mathit{\delta }}^{\mathrm{13}}\mathrm{C}={\displaystyle \frac{{\left(\frac{{}^{\mathrm{13}}\mathrm{C}}{{}^{\mathrm{12}}\mathrm{C}}\right)}_{\mathrm{sample}}}{{\left(\frac{{}^{\mathrm{13}}\mathrm{C}}{{}^{\mathrm{12}}\mathrm{C}}\right)}_{\mathrm{standard}}}}-\mathrm{1}\end{array}$$

For a detailed description of the δ¹³C–CH₄ measurements by GC-IRMS and technical details of the pre-concentration system, we refer the reader to previous studies by Comba et al. (2018) and Laukenmann et al. (2010).

2.8 Statistics

To test for significant differences in cell density, CH₄ formation, and CH₄ content between the treatments, two-way analysis of variance (ANOVA) (considering repeated measurements) and a post hoc test (Fisher least significant difference (LSD) test; alpha 5 %) were used.

3 Results

3.1 Algal growth and CH₄ formation

To investigate CH₄ production by algal cultures, incubations with ¹³C-labeled hydrogen carbon were applied as described in Sect. 2.3. The growth curves during incubation of the three algal species are presented in Fig. 3 (panels a, b, c). The initial cell densities were $\mathrm{26.9}\pm \mathrm{4.0}\times {\mathrm{10}}^{\mathrm{3}}$ cells mL⁻¹ for Chrysochromulina sp., $\mathrm{25.6}\pm \mathrm{1.2}\times {\mathrm{10}}^{\mathrm{3}}$ cells mL⁻¹ for P. globosa, and $\mathrm{17.5}\pm \mathrm{2.0}\times {\mathrm{10}}^{\mathrm{3}}$ cells mL⁻¹ for E. huxleyi. The exponential growth rate µ was highest for E. huxleyi (1.71±0.04 d⁻¹), i.e., 3 or 5 times higher than for P. globosa and Chrysochromulina sp. (with 0.33±0.08 d⁻¹ and 0.52±0.07 d⁻¹, respectively). The dotted lines in Fig. 1a, b, and c mark the time points of exponential growth. Maximum cell densities were lowest for Chrysochromulina sp. with $\mathrm{0.18}\pm \mathrm{0.01}\times {\mathrm{10}}^{\mathrm{6}}$ cells mL⁻¹, followed by E. huxleyi with $\mathrm{1.70}\pm \mathrm{0.09}\times {\mathrm{10}}^{\mathrm{6}}$ cells mL⁻¹, and highest for P. globosa with $\mathrm{1.77}\pm \mathrm{0.15}\times {\mathrm{10}}^{\mathrm{6}}$ cells mL⁻¹. Significant CH₄ formation was observed in all three cultures over the whole incubation period of 5 to 11 d (Fig. 3d, e, f), whereas no increase in CH₄ over time was observed in the control groups. For all species the increase in headspace CH₄ was significant (p≤0.05) at second time point of measurement and at all following time points (p≤0.001). At the end of the incubation period the amounts of produced CH₄ were 34.9±7.3, 99.3±8.2, and 45.0±3.1 ng for Chrysochromulina sp., P. globosa, and E. huxleyi, respectively. A linear correlation was found between the absolute number of cells and the amount of produced CH₄ of Chrysochromulina sp., P. globosa, and E. huxleyi (Fig. 3g, h, i).

Figure 3 Cell growth (a, b, c) and CH₄ production (d, e, f) in the course of time and the correlation between the total number of cells and produced CH₄ (g, h, i) from three algae species. Chrysochromulina sp. (a, d, g), P. globosa (b, e, h), and E. huxleyi (c, f, i). Please note that the cell numbers of Chrysochromulina sp. are presented in 10⁵ and those of P. globosa and E. huxleyi are in 10⁶. Mean values of six (Chrysochromulina sp., P. globosa) and three (E. huxleyi) replicated culture experiments are shown, and error bars mark the SD.

3.2 Stable carbon isotope values of CH₄ during incubation with ¹³C–hydrogen carbonate

Stable carbon isotope values of CH₄ (δ¹³CH₄ values) for Chrysochromulina sp. and P. globosa are presented in Fig. 4a, c. We observed the conversion of ¹³C carbon (provided by ¹³C–hydrogen carbonate) to ¹³CH₄ in cultures of both species, indicated by increasing δ¹³CH₄ values over time. Stable isotope values increased from initial atmospheric (laboratory air) levels of $-\mathrm{48.7}\pm \mathrm{0.3}$ ‰ and $-\mathrm{48.4}\pm \mathrm{0.10}$ ‰ up to $+\mathrm{30.1}\pm \mathrm{10.2}$ ‰ and $+\mathrm{245}\pm \mathrm{16}$ ‰ for Chrysochromulina sp. and P. globosa, respectively, whilst the δ¹³CH₄ values of the control groups (algae without ¹³C–hydrogen carbonate or ¹³C–hydrogen carbonate in medium without culture) did not change over time. The increase in δ¹³CH₄ values in the headspace CH₄ depended on the amount of released CH₄ that was added to the initial (atmospheric) background level. To calculate the δ¹³CH₄ values of the CH₄ source that raised the CH₄ quantity above background level, the Keeling plot method (Keeling, 1958; Pataki et al., 2003) was used (Fig. 4b, d). The calculated δ¹³CH₄ values of the CH₄ source were $+\mathrm{1300}\pm \mathrm{245}$ ‰ (Chrysochromulina sp.) and $+\mathrm{1511}\pm \mathrm{35}$ ‰ (P. globosa) and thus close to the theoretical calculated ¹³C value of the DIC (2014±331 ‰ ) resulting from the addition of ¹³C–hydrogen carbonate. Please note that ¹³C–hydrogen carbonate stable isotope labeling experiments with E. huxleyi were already performed by Lenhart et al. (2016) and were not repeated in this study. This is why δ¹³CH₄ values and the respective Keeling plot of E. huxleyi are not shown in Fig. 4.

Figure 4 δ¹³CH₄ values (a, c) and respective Keeling plots (b, d) from Chrysochromulina sp. (a, b) and P. globosa (c, d) after the addition of ${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$. Panels (a, c) show the δ¹³CH₄ values of three investigation groups (“culture + ${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$”, “culture”, and “${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$”), whereas each data point presented is the mean value of three replicated culture experiments with error bars showing SD. Panels (b, d) show the Keeling plots for the treatments “culture + ${\mathrm{H}}^{\mathrm{13}}{\mathrm{CO}}_{\mathrm{3}}^{-}$” from each replicated culture experiment (n₁, n₂, n₁), where f(0) refers to the ¹³C value of the CH₄ source.

3.3 CH₄ production and production potential

Since the experiment in the section above (isotope measurements) was not designed to obtain POC quotas, we conducted an additional experiment to estimate the CH₄ production rates of the three algal species. From an initial cell density of $\mathrm{22.5}\pm \mathrm{3.1}\times {\mathrm{10}}^{\mathrm{3}}$, $\mathrm{80.9}\pm \mathrm{11.5}\times {\mathrm{10}}^{\mathrm{3}}$, and $\mathrm{29.0}\pm \mathrm{5.5}\times {\mathrm{10}}^{\mathrm{3}}$ cells mL⁻¹ cultures were grown up to $\mathrm{37.0}\pm \mathrm{9.2}\times {\mathrm{10}}^{\mathrm{3}}$, $\mathrm{219}\pm \mathrm{24.1}\times {\mathrm{10}}^{\mathrm{3}}$, and $\mathrm{283}\pm \mathrm{15.6}\times {\mathrm{10}}^{\mathrm{3}}$ cells mL⁻¹ for Chrysochromulina sp., P. globosa, and E. huxleyi, respectively. These cell densities corresponded to the cell densities of the exponential growth phase obtained from the experiment in Sect. 3.1. The POC-normalized daily CH₄ production rate was highest in E. huxleyi, followed by P. globosa and Chrysochromulina sp. However, the cellular or POC-normalized daily production rates of the three algal species were of the same order of magnitude (Table 1). We calculated the CH₄ production potential (CH₄PP), which is the amount of CH₄ produced within a week of growth (Gafar et al., 2018), to translate the cellular production rates (µ × CH₄ cell⁻¹) of each species to the community level. The CH₄–PP was 2 orders of magnitude higher for E. huxleyi than the other two species. This is a consequence of the higher growth rate of E. huxleyi. We furthermore observed the oxygen concentrations during the light and dark periods to ensure oxic conditions. The measured oxygen concentrations were always saturated or supersaturated relative to equilibration with ambient air (Fig. S2).

Table 1 Growth rate, cellular POC, CH₄ production rates, and CH₄–PP_(7 d) of Chrysochromulina sp. (n=4), P. globosa (n=4), and E. huxleyi (n=4). Values are the mean of four replicated culture experiments with SD.

3.4 CH₄ formation from ¹³C-labeled methyl thiol ethers

The three methylated sulfur compounds MSO, DMSO, and DMS were tested for potential CH₄ formation in incubation experiments with E. huxleyi. The treatments were initiated in parallel from a batch culture by inoculating $\mathrm{17.5}\pm \mathrm{2.0}\times {\mathrm{10}}^{\mathrm{3}}$ cells mL⁻¹, and cultures were grown to final cell densities of $\mathrm{1.77}\pm \mathrm{0.08}\times {\mathrm{10}}^{\mathrm{6}}$ cells mL⁻¹ (Fig. 5a). Cell densities and CH₄ formation correlated in all treatments, while no difference in cell growth pattern or CH₄ formation was observed when isotope-labeled methyl thioether and sulfoxides were added to the culture (Fig. 5a, b, c). Differences between treatments were found in δ¹³CH₄ values of headspace CH₄. The initial δ¹³CH₄ value of headspace ($-\mathrm{47.9}\pm \mathrm{0.1}$ ‰, laboratory air) increased slightly over time in untreated cultures (without isotope treatment) to $-\mathrm{46.8}\pm \mathrm{0.3}$ ‰ (Fig. 6b). In contrast, experiments in which ¹³C₂–DMS, ¹³C₂–DMSO, and ¹³C–MSO were applied to cultures of E. huxleyi, δ¹³CH₄ values increased to $-\mathrm{31.0}\pm \mathrm{1.1}$ ‰, $-\mathrm{45.7}\pm \mathrm{0.1}$ ‰, and $+\mathrm{18.3}\pm \mathrm{7.7}$ ‰, respectively, over a time period of 6 d (Fig. 6a, b, c) and differed significantly from control groups (p<0.05). The results unambiguously show that a fraction of the ¹³C-labeled methyl groups of the added substances was converted to ¹³C–CH₄ in cultures of E. huxleyi. Much smaller changes in δ¹³CH₄ values were observed for controls of sterile filtered media to which only ¹³C₂–DMS and ¹³C–MSO were added ($-\mathrm{42.8}\pm \mathrm{1.7}$ ‰ and $-\mathrm{43.9}\pm \mathrm{0.2}$ ‰, respectively; Fig. 6a, c; day 6), whereas δ¹³CH₄ values did not change over time in the seawater controls (no addition of isotopic-labeled compounds) or in the seawater controls treated with ¹³C₂–DMSO (Fig. 6b). Based on the initial amount of ¹³C-labeled substance that was added to the cultures and the total amount of ¹³CH₄ at the end of the incubation period, 9.5±0.2 pmol (¹³C₂–DMS), 3.0±3.2 pmol (¹³C₂–DMSO), and 30.1±3.6 pmol (¹³C–MSO) of 8.5 µmol were converted to CH₄.

Figure 5 Cell growth (a), CH₄ production (b), and the relation between the total number of cells and produced CH₄ (c) from E. huxleyi treated with ¹³C₂–DMS, ¹³C₂–DMSO, and ¹³C–MSO or without any treatment. Mean values of three replicated culture experiments are shown, and error bars mark the SD.

Figure 6 δ¹³CH₄ values of headspace CH₄ in cultures of E. huxleyi supplemented with (a) ¹³C₂–DMS, (b) ¹³C₂–DMSO, and (c) ¹³C–MSO. Mean values of three replicated culture experiments are shown, and error bars mark the SD.

4 Discussion

Our results of CH₄ production and stable carbon isotope measurements provide unambiguous evidence that next to E. huxleyi (Lenhart et al., 2016) other widespread marine algal species, namely Chrysochromulina sp. and P. globosa, are involved in the production of CH₄ under oxic conditions at rates of 1.9±0.6 to 3.1±0.4 µg CH₄ g⁻¹ POC d⁻¹. The three investigated genera of marine phytoplankton have a worldwide distribution and they are representatives of the most widespread marine haptophytes (Schoemann et al., 2005; Thomsen, 1994; Brown and Yoder, 1994). The results indicate that CH₄ production could be a common process across marine haptophytes. We first discuss the stable isotopic evidence of CH₄ formation, the role of precursor compounds, and the likely mechanisms involved. Finally, we discuss the laboratory CH₄ production rates in relation to their potential significance in marine environments and provide a first rough estimation of how these production rates might contribute to CH₄ concentration in oxic surface waters previously reported in open-ocean algal blooms.

In cultures of Chrysochromulina sp. and P. globosa that were treated with ¹³C-labeled hydrogen carbonate, δ¹³CH₄ values increased with incubation time, clearly resulting from the conversion of ¹³C–hydrogen carbonate to ¹³CH₄. These results demonstrate that all three investigated algal species are instrumental in the production of CH₄ under oxic conditions (Fig. S1) and that hydrogen carbonate serves as a carbon source for ¹³CH₄. Our findings are in agreement with the stable isotope evidence of CH₄ production by E. huxleyi (Lenhart et al., 2016). However, we do not consider hydrogen carbonate to be the direct carbon precursor of CH₄. In a first step hydrogen carbonate and its isotope label are converted to CO₂ and subsequently fixed by algal primary production, forming POC. Therefore, we would expect a large fraction of the ¹³C label of the hydrogen carbonate ($+\mathrm{2014}\pm \mathrm{331}$ ‰) to be transferred to the POC towards the end of the experiment (when the volume-normalized POC content is highest). The experiments were started by the inoculation of cells from pre-cultures that were grown on DIC with natural ¹³C∕¹²C abundance (δ¹³C values ∼0 ‰). This means that during ongoing incubation the δ¹³C–POC value should get close to δ¹³C–DIC values, resulting from the addition of ¹³C–hydrogen carbonate when cultures grow in the new ¹³C-enriched medium. Consequently, the δ¹³C–POC values are considered to be somewhat lower than the theoretically calculated δ¹³C–DIC values ($+\mathrm{2014}\pm \mathrm{331}$ ‰) of the medium. Our assumptions are in line with the δ¹³CH₄ source signature values (averaged over 9 or 11 d, respectively) obtained via the Keeling plot method, which were $+\mathrm{1300}\pm \mathrm{245}$ ‰ and $+\mathrm{1511}\pm \mathrm{35}$ ‰ for Chrysochromulina sp. and P. globosa, respectively, and were thus somewhat lower than for the theoretical calculated ¹³C value of the DIC ($+\mathrm{2014}\pm \mathrm{331}$ ‰ ) resulting from the addition of ¹³C–hydrogen carbonate. Unfortunately, δ¹³C–DIC and δ¹³C–POC values could not be determined in our set of experiments to allow for more detailed calculations. However, our results clearly indicate that hydrogen carbonate is the principal inorganic carbon precursor of ¹³CH₄ produced in algae, but intermediate metabolites are likely to be formed from which CH₄ is released, possibly by cleavage of sulfur-bonded methyl groups of methyl thioethers and sulfoxides (Althoff et al., 2014; Lenhart et al., 2016; Benzing et al., 2017).

4.1 CH₄ formation from ¹³C-labeled methyl thioethers

4.1.1 Methyl thioethers are precursors of CH₄

Methyl thioethers and their sulfoxides are ubiquitous in marine environments as they are often produced by algae at substantial rates. It is also known that these compounds are metabolized in the three investigated algal species (Liss et al., 1994; Keller, 1989). Based on the addition of ¹³C₂–DMSO, ¹³C₂–DMS, and ¹³C–MSO, whereby only the sulfur-bonded methyl groups (–S–CH₃) were 99 % labeled with ¹³C, it was possible to clearly monitor ¹³CH₄ formation by stable carbon isotope measurements in cultures of E. huxleyi. The δ¹³CH₄ values increased over time significantly in ¹³C₂–DMS, ¹³C₂–DMSO, and ¹³C–MSO treated cultures, above the δ¹³CH₄ values of the control groups (Fig. 6a–c). The ¹³C labeling experiment showed that DMS, DMSO, and MSO are potentially important methyl precursors for CH₄, but the contribution of these compounds to the overall CH₄ production in cultures of E. huxleyi could not be determined in our experiments due to the complexity of the formation of these compounds in the algal cells. This can be illustrated by the following. The contribution of a substance to the total CH₄ released is the product of both the added ¹³C-labeled fraction (added to the water sample and uptake by the cells) and the internally formed fraction of these compounds (DMS, DMSO, and MSO), which will roughly show natural ¹³C abundance. Therefore, the stable isotope value of CH₄ will be diluted by the fraction of naturally formed methyl sulfur compounds in the algal cells, and thus the contribution of DMS, DMSO, and MSO to CH₄ formation cannot be estimated on the basis of their added amount alone. The ¹³CH₄ quantity from the conversion of added ¹³C-labeled substance contributed 0.03 % (¹³C₂–DMSO) up to 0.84 % (¹³C–MSO) to the overall released CH₄. However, even if the added ¹³C-labeled compounds might only explain ≤1 % of CH₄ formed by the algae, their overall contribution (including non-labeled sulfur compounds, which we are not able to measure) might provide a much larger fraction of the released CH₄. The intracellular DMS concentration can reach 1 mM (Sunda et al., 2002) in cells of E. huxleyi, while the concentration of added ¹³C₂–DMS was 0.01 mM in medium (final concentration). If intracellular ¹³C₂–DMS was in equilibrium with bulk seawater ¹³C₂–DMS and all CH₄ was produced from intracellular DMS, then the contribution of the ¹³C-labeled compound would be about 1 %. However, even if the biggest fraction of CH₄ in algae cultures was not released by the ¹³C-labeled substances, the significant increase in delta notation in ¹³C₂–DMS, ¹³C₂–DMSO, and ¹³C–MSO treated cultures above the δ¹³CH₄ values of the control groups demonstrates that ¹³C-labeled precursor substances were converted to CH₄ by algal cultures (Fig. 6a–c). This is also indicated when the absolute conversion quantities of ¹³C-labeled substance in algal cultures are considered: these were ca. 9 (¹³C₂–DMS), 3 (¹³C₂–DMSO), and 30 (¹³C–MSO) times higher than in seawater control groups. Hence, the stable isotope labeling approach should be considered as a proof of concept that methyl groups of all tested substance serve as precursor compounds of CH₄. These isotope labeling results are also in good agreement with recent results from laboratory experiments in which ¹³C–MET was added to cultures of E. huxleyi (Lenhart et al., 2016).

In addition, we also found an indication for a purely chemical CH₄ formation pathway from control samples using sterile seawater and the addition of either ¹³C₂–DMS or ¹³C–MSO. The ¹³C₂–DMS-spiked seawater group and the ¹³C₂–DMS-spiked algae group are very close to each other up to day 2 (see Figs. 5a and 6a). For this time period, it can be assumed that the chemical conversion has taken place in both samples to the same extent, since the samples are relatively similar because the algal cell density is only 5 % (day 2) of the final cell density. However, the following days (day 3 to day 6), when algal cell density increased drastically, the δ¹³CH₄ values of the algae cultures also increased significantly compared with δ¹³CH₄ values of the seawater. This clearly indicates that the conversion of ¹³C₂–DMS to CH₄ increases with increasing cell counts. However, the relatively slight increase in δ¹³CH₄ values in the control samples (Fig. 6a, c) indicates that this is only a minor pathway. The CH₄ conversion from ¹³C–DMS and ¹³C–MSO in seawater was approximately 3- and 30-fold lower than in the corresponding treatments with algae and only becomes obvious when applying very sensitive stable isotope labeling experiments. A similar observation was already made by Lenhart et al. (2016) when applying ¹³C–MET in seawater. However, this observation might be in agreement with previously findings by Zhang et al. (2015), who described a photochemically and CDOM-related conversion of DMS to CH₄ in oxygenated natural seawater.

4.1.2 Potential mechanism of CH ₄ formation from thioethers

The CH₄ formation from thioethers (MET, DMS) and their corresponding sulfoxides (MSO, DMSO) might be catalyzed by nonheme iron-oxo (IV), thus forming methyl radicals (• CH₃) from homolytically broken sulfur methyl bounds (R–CH₃), leading to CH₄ under oxidative conditions (Althoff et al., 2014; Benzing et al., 2017). The tested compounds are found in high cellular concentrations in E. huxleyi, Chrysochromulina sp., and P. globosa, and nonheme iron-oxo (IV) has been identified as an active intermediate in the catalytic cycles of a number of biological enzymatic systems (Hohenberger et al., 2012). Therefore, the postulated reaction might be a likely pathway for CH₄ production in investigated algal species. Furthermore, DMS and DMSO were described to be part of an antioxidant system as these compounds can readily scavenge hydroxyl radicals in cells of E. huxleyi (Sunda et al., 2002). Furthermore, CH₄ is released via a methyl radical that is subtracted from DMSO when hydroxyl radicals are being scavenged – and accordingly DMS after its sulfoxidation (Herscu-Kluska et al., 2008). Since MET and MSO have similar functional groups to DMS and DMSO, respectively, it was proposed that the reaction described above is taking place analogously for these compounds (Bruhn et al., 2012; Lenhart et al., 2015a). Consequently, the CH₄ formation in investigated algal species might be a response of oxidative stress that forms hydroxyl radicals or other reactive oxygen species (ROS), which in turn might react with the applied methylated sulfur compounds, generating methyl radicals and eventually CH₄.

The algal metabolites DMSP, DMS, and DMSO are ubiquitous in marine surface layers and nanomolar concentrations were found in blooms of Chrysochromulina sp., P. globosa, and E. huxleyi. Several field studies showed that these compounds are linked to CH₄ formation in seawater (Zindler et al., 2013; Damm et al., 2008; Florez-Leiva et al., 2013). Various authors have proposed that DMSP and its degradation products DMSO and DMS are used by methylotrophic methanogenic archaea, inhabiting anoxic microsites, as substrates for methanogenesis. In addition it was reported that, if nitrogen is limited but phosphorus is replete, marine bacteria might also use DMSP as a carbon source, thereby releasing CH₄ (Damm et al., 2010). One scenario which we cannot rule out would be the production of CH₄ precursors by algae and the usage of these precursors by bacteria to produce CH₄. While we think that this is less likely than CH₄ production by algae alone, it would, if true, show that bacteria need algae-produced precursors to produce CH₄. The latter scenario would be relevant in the field because algae coexist with bacteria in the oceans. Therefore, bacteria might be involved in the CH₄ production process in our cultures, but even if they were, they would still depend on algal growth. For further discussion of a potential contribution of heterotrophs and/or methanogenic archaea, see the Supplement (S3). The correlations we describe in the Supplement clearly show that CH₄ production depends on algal growth. It is therefore highly unlikely that bacteria are solely responsible for CH₄ production in our cultures.

4.2 POC-normalized production

For all three algal species significant correlations between CH₄ mass and cell density were found (r²>0.95 for all species; Fig. 3g, h, i), suggesting that CH₄ formation occurred over the entire growth curve. However, since CH₄ production can only be determined in the exponential phase (Langer et al., 2013) we additionally ran dilute batch cultures to determine CH₄ production. All three species displayed similar CH₄ production ranging from 1.9±0.6 to 3.1±0.4 µg CH₄ g⁻¹ POC d⁻¹, with Chrysochromulina sp. and E. huxleyi showing the lowest and highest rates, respectively. The CH₄ production for E. huxleyi was found to be 2-fold higher than rates reported for the same strain and comparable culture conditions by Lenhart et al. (2016) (0.7 µg CH₄ g⁻¹ POC d⁻¹). The lower production reported by Lenhart et al. (2016) may be explained by the fact that CH₄ production was not obtained from exponentially growing cultures. We also compared the cellular CH₄ production rates of E. huxleyi reported by Scranton (1977) with those of our study. Scranton (1977) reported a production rate of $\mathrm{2}\times {\mathrm{10}}^{-\mathrm{10}}$ nmol CH₄ cell⁻¹ h⁻¹. This value is close to the production rate of $\mathrm{1.6}\times {\mathrm{10}}^{-\mathrm{10}}$ nmol CH₄ cell⁻¹ h⁻¹ in our study. Scranton (1977) concluded from observed CH₄ production rates in laboratory experiments that natural populations might be adequate to support the widespread supersaturation of CH₄ observed in the open ocean. However, we suggest that the CH₄ production of various algae might differ substantially under changing environmental conditions, as already shown for terrestrial plants (Abdulmajeed and Qaderi, 2017; Martel and Qaderi, 2017). Moreover, the cellular concentrations of potential precursor compounds such as methylated sulfur compounds might vary greatly between species and cultures. The investigated algal species can reach millimolar intracellular concentrations of DMS and DMSP (Sunda et al., 2002; Liss et al., 1994; Keller, 1989), and even if the conversion rate of methylated sulfur compounds to CH₄ in algal cells might be low, they could be sufficient to explain a substantial fraction of the CH₄ production rates by marine algae.

4.3 Implication for the marine environment and algal blooms

In general, the distribution of chlorophyll has not shown a consistent correlation with CH₄ distributions in field studies. There are studies in which no correlation was observed (e.g., Lamontagne et al., 1975; Foster et al., 2006; Watanabe et al., 1995) or a correlation was found within a few depth profiles (Burke et al., 1983; Brooks et al., 1981). Many field measurements in oxygenated surface waters in marine and limnic environments have shown examples of elevated CH₄ concentrations spatially related to phytoplankton occurrence (e.g., Conrad and Seifer, 1988; Owens et al., 1991; Oudot et al., 2002; Damm et al., 2008; Grossart, et al., 2011; Weller et al., 2013; Zindler et al., 2013; Tang et al., 2014; Bogard et al., 2014; Rakowski et al., 2015). Taken together these studies suggest that phytoplankton is not the sole source of CH₄ in oxygenated surface waters, but importantly, they also suggest that phytoplankton is one of the sources of CH₄. We therefore compared the CH₄ production rates of our cultures with two field studies for the Pacific Ocean (Weller et al., 2013) and the Baltic Sea (Schmale et al., 2018) to evaluate the potential relevance of algal CH₄ production. It was estimated that the gross CH₄ production in a southwest Pacific Ocean mesoscale eddy is 40–58 pmol CH₄ L⁻¹ d⁻¹ (Weller et al., 2013). Using reported phytoplankton cell densities (1.7×10⁸ to 2.9×10⁸ cells L⁻¹, Weller et al., 2013), we calculated a maximal cellular production of 5.5 ag CH₄ cell⁻¹ d⁻¹ for this eddy. The species investigated in this study showed ca. 3–11 times higher cellular production (Table 1). Hence, each of the three haptophyte algae studied here could account for the CH₄ production reported by Weller et al. (2013).

Schmale et al. (2018) reported CH₄ enrichments that were observed during summer in the upper water column of the Gotland Basin, central Baltic Sea. Furthermore, they found that zooplankton is one but not the only CH₄ source in the oxygenated upper waters. While the authors ruled out a major contribution of algae to the observed sub-thermocline CH₄ enrichment because of the low sub-thermocline phytoplankton biomass, they considered primary-production-associated CH₄ formation to be one likely source in the phytoplankton-rich mixed layer. The average phytoplankton carbon biomass of the mixed layer was approximately 600 µg L⁻¹ (averaged from Fig. 9 in Schmale et al., 2018). For the reported average net CH₄ production rate in the mixed layer (95 pmol CH₄ L⁻¹ d⁻¹), we calculated that a production rate of 2.5 µg g⁻¹ POC d⁻¹ is required if the CH₄ is produced by the algal biomass. This rate would be within the range of CH₄ production rates observed in our study. These calculations should be considered as a first rough estimate to assess whether the CH₄ production rates of laboratory-grown cultures can significantly contribute to CH₄ supersaturation associated with phytoplankton. We did not distinguish between species and did not take into account environmental factors or the complexity of microbiological communities.

Judging from cellular production, the species studied here are of similar importance for oceanic CH₄ production in biogeochemical terms. Regarding the highest cellular production, with that of E. huxleyi as 100 %, P. globosa produces 27 % and Chrysochromulina sp. 71 % (Table 1). However, several recent studies have emphasized that production potential (PP), as opposed to cellular production, is a biogeochemically meaningful parameter (Gafar et al., 2018; Marra, 2002; Schlüter et al., 2014; Kottmeier et al., 2016). The concept of production potential goes back at least to the first half of the 20th century (Clarke et al., 1946). Briefly, the production potential of substance X is the amount of X which a phytoplankton community or culture produces in a given time. For details, see Sect. 2 and the references above. Cellular production, by contrast, is the rate of production of X of a single cell, and therefore the cellular production is ill-qualified to express community-level production. We calculated the CH₄–PP (Sect. 2) for our three species, and when the one of E. huxleyi is considered to be 100 %, P. globosa has a CH₄–PP of 0.9 % and Chrysochromulina sp. 0.8 % (Table 1). In terms of CH₄ production in the field, therefore, E. huxleyi outperforms the other two haptophytes by 2 orders of magnitude. It can be concluded that the CH₄–PP under given environmental conditions is species-specific, and therefore community composition will have an influence on algal sea surface water CH₄ production.

It can be hypothesized that changing environmental conditions might drastically affect algal CH₄ production, which has to be taken into account when calculating annual averages. The effect of dominant environmental parameters such as light intensity and temperature on algal CH₄ production will therefore be the subject matter of future studies.