New index of organic mass enrichment in sea spray aerosols linked with senescent status in marine phytoplankton

Chemical properties of submicrometer aerosols

Figure 1a presents the time series of the mass concentrations of OC and water-soluble OC (WSOC) in the submicrometer aerosol particles during the cruise. Overall, the patterns of the temporal OC and WSOC variations were similar to each other (r = 0.632, p < 0.001), showing greater variability throughout the entire cruise. The concentrations of OC exceeded 3000 ngC m⁻³ in some cases. The average concentrations of OC and WSOC were 2278 ± 2067 ngC m⁻³ and 717 ± 440 ngC m⁻³, respectively. These concentration levels are 2–3 times higher than those in submicrometer particles observed during the post-bloom period in the western subarctic Pacific^19,20, likely because the current study region includes the coastal region.

Temporal variations of glucose and methanesulfonic acid (MSA) in the submicrometer aerosols are shown in Fig. 1b. Glucose is used here as a tracer for primary marine aerosols, as SSAs contain a substantial amount of monosaccharides, including glucose^21,22. MSA is considered to be either produced by gas-phase MSA directly scavenged by aerosols or rapidly produced in the aqueous phase from scavenged dimethylsulfoxide (DMSO) and methanesulfinic acid (MSIA)²³, particularly under conditions with high relative humidity typical of the marine boundary layer (MBL). Throughout the cruise, the temporal variations of glucose and MSA were generally similar to those of OC and WSOC on a timescale of a few days. This similarity in the temporal trends of the mass concentrations suggests that the observed OC and WSOC in the submicrometer aerosols were significantly affected by marine biogenic sources. It is noteworthy that the timing of increases in the concentrations of glucose relative to those of MSA sometimes differed on a half-day to a day scale, likely due to the different contributions of primary and secondary sources to the observed aerosol mass in each aerosol sample.

Chemical and biological properties of surface seawater

Figure 1c shows the concentrations of dissolved OC (DOC) and particulate OC (POC) in surface seawater at each sampling station corresponding to the aerosol sampling locations. Surface seawater DOC is defined as organic matter that penetrates through a 0.22-μm Durapore filter and contains particulate matter including viruses and colloids²⁴. DOC concentrations during the cruise varied within a narrow range (59.4–72.1 μM C), with an average of 65.9 ± 3.7 μM C. During the same expedition, fluorescence intensities of protein-like component in DOC determined by excitation-emission matrix and parallel factor analysis showed the higher level in the sea surface, whose spatial distribution was similar to that of Chl a²⁴. POC in surface seawater is operationally defined here as organic matter retained on a pre-combusted Whatman GF/F glass-fiber filter (nominal pore size of 0.7 μm)¹⁸, which contains a mixture of living planktonic organisms and detritus. POC concentrations in the surface seawater samples ranged from 5.0 to 23.7 μM C with an average of 12.0 ± 5.2 μM C. On average, DOC dominated TOC (= POC + DOC), which accounted for 85 ± 5% of TOC during the cruise.

Chl a concentrations ranged from 0.27 to 2.49 mg m⁻³ with an average of 0.88 ± 0.70 mg m⁻³ in this study. Concentrations of Chl a and POC in surface seawater showed a significant positive correlation (R² = 0.60) during the sampling period (Fig. 1c,d). At all the seawater sampling stations, diatoms were predominant, accounting for 54–96% of the total Chl a, which was determined by multiple regression analysis based on diagnostic pigment signatures²⁵. Furthermore, Yoshida et al.²⁵ reported that Thalassiosira generally dominated the diatom assemblages in seawater during the cruise. Chllide a is a breakdown product of Chl a by the enzyme chlorophyllase, which has been used as a biomarker for senescent algal cells, particularly in diatoms^17,26. For the bulk surface seawater in this study, the Chllide a concentrations ranged from 0.02 to 0.55 mg m⁻³ with an average of 0.09 ± 0.14 mg m⁻³, an order of magnitude lower than those of Chl a (Fig. 1d). The average concentration ratio of Chllide a/Chl a (9.6%) was well within the range (< 12%) in the pre-bloom phase reported in a similar oceanic region¹⁷. It should be noted that even under pre-bloom conditions, senescent phytoplankton cells could occur, because resource (light and/or nutrient) requirements differ among species (e.g., Sarthou et al.²⁷), and lower temperatures reduced the photosynthetic capability of phytoplankton in coastal Oyashio waters during the cruise²⁵. Although the fraction of Chllide a relative to Chl a was small, the current results clearly suggest that senescent cells existed even under the pre-bloom condition.

Here, we used an indicator (I_senes) for the relative abundance of senescent algal cells to the total phytoplankton cells, which is defined as follows:

where [Chl a] and [Chllide a] represent the concentrations of Chl a and Chllide a in the surface seawater, respectively¹⁷. Suzuki et al.¹⁷ showed that an index of diatom bloom development is positively correlated with I_senes regardless of the oceanic region in the western subarctic Pacific, demonstrating the validity of this I_senes. In the present study, variations of I_senes were generally similar to those of Chl a, but in some cases, these parameters showed an opposite trend on a timescale of a day (e.g., Mar. 11, 16, 18, and 23; Fig. 1d).

Figure 2 presents the relationship between the DOC/TOC ratio and I_senes in the surface seawater samples obtained during the entire cruise. In general, the DOC/TOC tended to increase with increasing I_senes values. This relationship supports a possible coupling between DOC and senescent status in marine phytoplankton. The relative increase in DOC with increasing I_senes is partly attributable to an increase in extracellular organic matter produced by senescent phytoplankton. Indeed, Nosaka et al.¹⁸ observed higher ratios of DOC/TOC produced by phytoplankton during the post-bloom phase in the Oyashio waters of the western North Pacific. They suggested that the excretion of DOC, such as acid polysaccharides by diatoms, can produce transparent exopolymer particles (TEP) in seawater over this oceanic region.

Organic mass enrichment in sea spray aerosols versus a biomarker for senescent algal cells

To investigate OM enrichment in SSA for senescent status in sea surface phytoplankton, submicrometer aerosol data representing SSA characteristics need to be selected from the ambient aerosol measurements. Characterization of SSAs in ambient air requires sampling and analytical methods that isolate these particles from being affected by preexisting or transported ambient gases and particles. A method using stable carbon isotope ratios (δ¹³C) combined with molecular markers provides robust tools to determine the contributions of marine and terrestrial sources to OC in marine aerosols^{22,28,29,30,31}. Miyazaki et al.³¹ specifically selected aerosol data with characteristics of SSA observed during the same cruise as in this study. Their method for the selection of SSA samples was based on the δ¹³C of TC (δ¹³C_TC) and water-soluble OC (WSOC; δ¹³C_WSOC) together with local wind speeds. They showed a significant positive correlation between WSOC and glucose in the SSA samples, supporting the SSA’s nascent characteristics. The SSA sample is defined here as that showing the δ¹³C_TC and δ¹³C_WSOC values higher than − 22‰ under conditions of local wind speeds of > 5 m s⁻¹. In this study, six samples were designate as SSA samples. For the selected SSA samples, all δ¹³C_TC values were within the values of marine sources with a negligible contribution of elemental carbon (EC; < 0.02 μg C m⁻³, see Fig. 1 of Miyazaki et al.³¹). Therefore, δ¹³C_TC can be assumed here to represent the δ¹³C value of OC. It is noted that seawater temperature may possibly affect the δ¹³C values in aerosols due to the isotope equilibrium fractionation linked to photosynthesis of phytoplankton³². However, the relation between the δ¹³C in SSA samples and surface sea temperature (SST) was not distinctly observed during aerosol sampling, which is partly attributed to the small variation in the average SST (~ 0.6–1.4 °C) during SSA sampling in this study³¹.

To explore the correlation between OM enrichment in the SSA samples and phytoplankton biomass in terms of Chl a or their senescent status, Fig. 3 presents the OC/Na⁺ and WSOC/Na⁺ ratios in the SSA as functions of the Chl a concentrations and I_senes. The term “enrichment” here refers to the amount of OC relative to Na⁺ in the SSA samples in this study. For the 12-h aerosol sampling, the OC/Na⁺ and WSOC/Na⁺ generally showed larger values with increasing the Chl a concentrations, although their correlations were not evident (r < 0.306, p > 0.05; Fig. 3a). The results were in agreement with the previous study that showed a correlation coefficient between satellite-derived Chl a and OM enrichment in sea salt particles on a timescale of a day, was lower than those on a weekly and monthly time scale¹⁴. In contrast, it is apparent that the OC/Na⁺ and WSOC/Na⁺ ratios showed a significant positive relationship with I_senes with r of 0.768 (p < 0.01) and 0.721 (p < 0.01), respectively (Fig. 3b). This result suggests that the OM enrichment in SSA is closely linked with the abundance of senescent algal cells relative to the total phytoplankton cells on a timescale of a day. Our results also support the concept that the production of OM in SSA is likely linked to senescent algal cells and/or cell lysis. It is noted that the OC/Na⁺ ratios observed in this study were generally larger than the typical range (0.1–2.0) previously reported for submicrometer marine primary OA³³. This is probably because the ratio in the coastal region as observed in this study is generally larger than that in the open ocean²¹.

Previous studies suggest that OM enrichment in SSA is driven by insoluble colloids and nanogels³⁴, forming surface-active organic materials. These organic materials are sufficiently small for inclusion as the operationally defined DOC, which contains viruses and cellular fragments owing to cell death or cell burst induced by a viral infection. Gel-like particles are composed of macromolecules and colloids. These organic materials are produced from surface-active polysaccharides and proteinaceous matter, which are ubiquitous at the ocean surface. In the present study, the observed OM in SSA, particularly WSOC, showed a significant correlation with glucose (r = 0.931, p < 0.001), a decomposed product of polysaccharides. The current results indicate that senescent status is particularly sensitive to OM enrichment in SSA during the pre-bloom phase on a daily timescale.

Hygroscopicity parameters of sea spray aerosols and senescent status in marine phytoplankton

With the water-extracted SSA samples defined here, the CCN activity was measured as the hygroscopicity parameter κ by nebulization of the filter extracts. Figure 4 displays κ in the SSA samples as a function of the corresponding I_senes in surface seawater. The κ values ranged between 0.50 and 0.65, with an overall average of 0.57 ± 0.06. These κ values are at the upper end of the range of values for ambient marine aerosol particles obtained by in-situ field measurements in high latitude oceanic areas (typically 0.1–0.6)^35,36. Given that the aerosol from the filter extracts is completely water-soluble, this is to be expected. Although the number of available data is limited, the κ value tends to be smaller with increasing I_senes (Fig. 4). This implies that the reduction of hygroscopicity in SSA is associated with the aging process of phytoplankton cells and the subsequent increase of OM in SSA.

The current result is similar to previous studies showing a decrease in the hygroscopicity parameter by 9–37% when phytoplankton exudates or DOC-rich algae cultures were added to seawater in a laboratory experiment³⁷. Furthermore, in that previous experiment, CCN activity has been found to depend on the type of algal exudate using a seawater proxy³⁷, where the experiment with diatomaceous exudate showed more significant effects on the critical supersaturation than the experiments with nanoplankton exudate. Fuentes et al.³⁷ found that the reduction in CCN activity was taxon-specific, with the DOM released by diatoms causing a greater reduction in CCN activity than that released by prymnesiophytes. Specifically, independent cultures of Thalassiosira Rotula, Chaetoceros sp., Emiliania Huxleyi, and Phaeocystis cf. Globosa in artificial seawater showed κ values of 0.95–1.21, which were lower than those (1.3–1.5) for the artificial seawater alone in their study. In wave channel experiments involving the heterotrophic bacterium Alteromonas spp. and the green microalga Dunaliella Tertiolecta, Collins et al.³⁸ reported κ values as low as 0.12. In the current study, κ of OM (κ_OM) for the SSA samples was calculated to be 0.44 ± 0.12 on the basis of the chemical fraction in the SSA samples (see methods section). The κ values shown in this study are within a range of the above values derived by the laboratory experiments. This study points to a reduction in the SSA hygroscopicity with increasing senescent status of diatom, suggested by the field measurements for the first time.

To summarize, the current study provides a new application of the index using concentrations of Chllide a, which is a biomarker for senescent algal cells (particularly diatoms), combined with Chl a concentration to represent the amount of OM in SSAs in the pre-bloom phase. This index should be further examined if it applies to the different stages of the bloom, namely phytoplankton bloom and post-bloom period, in future studies. Most bloom-forming diatoms show the chorophyllase activity converting from Chl a to Chllide a, whereas some algal groups such as the dinoflagellate Gymnodinium species, which sometimes bloom in coastal waters, do not possess the enzyme activity^39,40. Therefore, the algal senescent index would be particularly useful during diatom bloom. Because diatoms are distributed worldwide in surface waters⁴¹, the new index has the potential to be a robust marker of the OM amount in SSA.

It should be noted that satellite remote sensing cannot distinguish Chllide a from Chl a, because optical properties of these two pigments are almost identical⁴². Instead, field data of Chllide a and Chl a concentrations over a variety of oceanic regions are publicly available (e.g., NASA bio-Optical Marine Algorithm Dataset (NOMAD)⁴³, https://seabass.gsfc.nasa.gov/wiki/NOMAD). These datasets can be implemented in the climate models. In order to develop realistic parametrizations of the organic enrichment in SSA and subsequent emission to the atmosphere, further measurements of surface seawater and nascent SSA properties are required to confirm the results presented here.

Source: Ecology - nature.com