Volume 48, Issue 19 e2021GL093738
Research Letter
Free Access

Contribution of Marine Phytoplankton and Bacteria to Alkalinity: An Uncharacterized Component

Chang-Ho Lee

Chang-Ho Lee

Division of Environmental Science and Engineering, Pohang University of Science and Technology, Pohang, Korea

Search for more papers by this author
Kitack Lee

Corresponding Author

Kitack Lee

Division of Environmental Science and Engineering, Pohang University of Science and Technology, Pohang, Korea

Institute for Convergence Research and Education in Advanced Technology, Yonsei University, Seoul, Korea

Correspondence to:

K. Lee,

[email protected]

Search for more papers by this author
Young Ho Ko

Young Ho Ko

OJEong Resilience Institute, Korea University, Seoul, Korea

Search for more papers by this author
Joon-Soo Lee

Joon-Soo Lee

Ocean Climate and Ecology Research Division, National Institute of Fisheries Science, Busan, Korea

Search for more papers by this author
First published: 21 September 2021
Citations: 4


The contributions of phytoplankton and bacteria cells to alkalinity (AT) were measured in seawater samples obtained from 205 locations including the East Sea, the North Pacific Ocean, the Bering Sea, the Chukchi Sea, and the Arctic Ocean. We attributed the differences in AT values measured for unfiltered versus filtered samples to AT components contributed by phytoplankton (retained on a 0.7 μm filter) and by phytoplankton and bacteria combined (AT−BIO; retained on a 0.45 μm filter). The AT−BIO values reached 10–19 μmol kg−1 in the East Sea and the North Pacific Ocean, and progressively decreased to a level of 1 μmol kg−1 with distance toward the Arctic Ocean. The study shows that the AT−BIO values are non-negligible in coastal and open ocean environments and need to be considered when assessing the accuracy of carbon parameters calculated using the thermodynamic models that use measured AT as an input parameter.

Key Points

  • Phytoplankton and bacteria cells contributed to alkalinity during the titration of seawater

  • The biological components in alkalinity were non-negligible in coastal and open ocean environments

  • The biological components in alkalinity varied markedly depending on phytoplankton species in seawater

Plain Language Summary

In seawater thermodynamic calculations involving measured alkalinity, estimates of alkalinity contributed by bicarbonate and carbonate ions only are needed, and are made by correcting measured alkalinity values for contributions from borate and hydroxide ions, and from other minor chemical species. Among those minor components, contributions from phytoplankton and bacteria cells are an uncharacterized component of measured alkalinity. The present study shows that the contribution of particulate organic matter (mostly phytoplankton and bacteria) to measured alkalinity is significant in ocean environments. The common practice of making alkalinity measurements using unfiltered seawater in field studies could lead to overestimation of carbonate alkalinity when converting total alkalinity to carbonate alkalinity, because most total alkalinity data have been assumed to have negligible biological components. This overestimation could subsequently lead to errors in the calculated parameters. Thus the biological component in the total alkalinity needs to be taken into account when predicting carbon parameters using the thermodynamic models that use measured alkalinity as an input parameter.

1 Introduction

Seawater alkalinity (AT) equates to the total amount of all dissolved inorganic components that react with hydrogen ions during seawater titration by hydrochloric acids, to the pH point at which all carbonate and bicarbonate ions eventually become carbonic acids. As the evaluation of titration data for determination of seawater AT hinges on accurate data on the dissociated forms of salts and the acid-base reactions in seawater, the definition of AT has evolved in parallel with an improvement of our understanding of ocean biogeochemical processes, including redox reactions by organisms and interactions between minerals and aqueous solutions (e.g., chemical weathering and the precipitation and dissolution of CaCO3; Dickson, 1992; Hartmann et al., 2013; Wolf-Gladrow et al., 2007). The definition of AT widely used by the scientific community was proposed by Dickson (1981). It equates (Equation 1) to the amount of titrant hydrogen ions equivalent to the excess of conjugate bases dissociated from weak acids having K ≤ 10−4.5 over acids produced from weak bases having K > 10−4.5 in one kilogram of solution at zero ionic strength at 25°C:
where [H+]F is the concentration of free hydrogen ions. Carbonate species ([HCO3] + 2[CO32−]) make by far the greatest contribution to AT, followed by borate ions ([B(OH)4]) and other minor components (e.g., [OH] + [HPO42−] + 2[PO43−] + [SiO(OH)3]). The magnitude of each component of AT varies with location, and the other uncharacterized chemical species that contribute to AT are represented in the last term in Equation 1.

Among the uncharacterized species, the contribution of dissolved organic matter (DOM) to AT (AT−DOM) was initially neglected or assumed to be zero; however, its importance has recently been recognized, but studied only in limited environments (Sharp & Byrne, 2020). Since Goldman and Brewer (1980) suggested that organic acids could contribute to AT, subsequent studies in estuaries (Cai et al., 1998), coastal oceans (Hernández-Ayon et al., 2007; Ko et al., 2016), regions having high inputs of terrestrial matter (Kuliński et al., 2014), and those involving culture experiments (Hernández-Ayon et al., 2007; Kim & Lee, 2009) have shown that measured AT values deviate from those calculated from other CO2 parameters using the optimal thermodynamic model for the ocean CO2 system. More compellingly, with respect to the dissolved organic carbon (DOC) concentration, the differences between measured AT versus calculated AT have been strongly associated in some areas (R2 > 0.9; Kuliński et al., 2014) but weakly associated in others (R2 < 0.25; Song et al., 2020). The lack of a consistent association between AT−DOM and DOC indicates that either not all DOM is involved in AT or all DOM do not equally contribute to AT. In some cases, the AT−DOM values exceeded the combined contribution from all minor components (e.g., [OH] + [HPO42−] + 2[PO43−] + [SiO(OH)3]), and in rare cases they even exceeded the contribution of borate ions to AT (AB = 50–80 μmol kg−1; Hernández-Ayon et al., 2007; Song et al., 2020).

Another component of AT yet to be fully characterized is phytoplankton and bacteria cells. In a bay of Korea the surfaces of phytoplankton and bacteria cells contributed to AT (Kim et al., 2006). During seawater titration these cell surfaces as proton acceptors are closely associated with their macromolecular components (e.g., phospholipids and lipoproteins on microalgae; peptidoglycans on bacteria cells), which tend to be rich in hydroxyl, carboxyl, amine, or phosphate groups (Natarajan, 2018; Yuan et al., 2019). These functional groups on the cell surfaces can respond to changes in seawater pH through adsorption of ions (Poortinga et al., 2002), being protonated at low pH or deprotonated at high pH levels. Hence, in weakly alkaline seawater these microbes tend to have net negative charges on their cell surfaces, so their proton sorption during seawater titration directly contributes to AT (Middelburg et al., 2020).

To our knowledge, no previous reports have examined how the AT component associated with phytoplankton and bacteria cells varies among different open ocean environments. The primary aim of the present study was therefore to assess the contribution of phytoplankton and bacteria cells (AT−BIO) to AT in diverse oceanic environments.

Another aim of this study was to evaluate the effect of accounting for AT−BIO when evaluating the accuracy of calculated carbon parameters using seawater carbonate thermodynamic models. Specifically, in the thermodynamic calculations involving AT as an input parameter, the AT needs to be converted to carbonate alkalinity (AC = [HCO3] + 2[CO32−]) by subtracting the non-carbonate contributions (largely from borate ions). Performing AT measurements using unfiltered seawater, a common practice in field studies, could lead to an overestimation of AC, because most AT data collected to date have been assumed to have a negligible AT−BIO component. This overestimation could lead to inaccuracies in the calculated CO2 parameters.

2 Methods and Materials

2.1 Sampling Locations

A total of 205 surface samples were collected in 4-L polycarbonate bottles from the underway clean seawater line (sampling locations are indicated by colored circles in Figure 1a). To obtain depth profiles, a total of 24 discrete samples were collected in 1-L high density polyethylene bottles using a rosette sampler (SeaBird Electronics, SBE 911 plus) at two sites in the Arctic Ocean (cross symbols in Figure 1a).

Details are in the caption following the image

(a) Surface seawater sampling locations (colored circles) in the East Sea (N = 20), the North Pacific Ocean (N = 40), the Bering Sea (N = 44), the Chukchi Sea (N = 28), and the Arctic Ocean (N = 73) including the East Siberian Sea. The sizes and colors of circles indicate the magnitude of contributions of all biogenic particles to AT (AT−BIO). AT−BIO values <1.1 μmol kg−1 (measurement precision of AT) are shown as red dots. Two sampling sites for depth profiles in the Arctic Ocean are shown by black crosses. The AT−BIO depth profiles are shown in the inset, where the dashed line indicates the mean AT−BIO (0.7 μmol kg−1). The color gradation in the background shows the chlorophyll-a concentration (obtained from Moderate resolution Imaging Spectroradiometer (MODIS) Aqua) averaged for the study period. (b) AT−BIO, AT−PLANKTON, AT−BACTERIA, and AT−PIC for the five basins studied. The middle, lower, and upper horizontal lines in the boxes indicate the median, and the first and third quartiles of each component, respectively. The lower and upper whiskers show the minimum and maximum values, and the colored dots show the outliers.

Sampling was carried out on board the R/V ARAON from 17 July 2020 to 15 September 2020. The sampling locations were approximately evenly spaced along a transect from the East Sea to the North Pacific Ocean, the Bering Sea, the Chukchi Sea, and the Arctic Ocean (Figure 1a). Half of the samples were collected on the route to the Arctic Ocean, and the other half were sampled on the return route to Korea. Each of the five basins sampled has distinct characteristics in phytoplankton biomass and species composition. High biomass in the East Sea is primarily sustained by a high nutrient input from the deep basin through the ventilation and additionally increasing input of anthropogenic nutrient from the continent (Lee et al., 2011; Park et al., 2006), whereas the North Pacific Ocean receives a high nutrient flux from the intrusion of cold dense Okhotsk waters, and thus phytoplankton in this basin are productive (Kim, 2012). Coccolithophores (which form CaCO3 coccolith plates) are abundant in the Bering Sea, and their presence leads to non-negligible particulate inorganic carbon concentrations (Broerse et al., 2003; Ladd et al., 2018). The concentration of particulate organic carbon generally decreases with distance northwards to the Arctic Ocean.

2.2 Filtration of Organic and Inorganic Particulates for AT Measurement

Filtration of seawater is critical in enabling accurate quantification of the contributions of phytoplankton, and bacteria, and biogenic CaCO3 particles to AT. Prior to filtration, filter funnels and flasks were rinsed with the seawater sample to be filtered. The seawater sample was passed through filters having pore sizes of 0.7 μm (glass microfiber filter; Whatman; combusted at 450°C for 5 h prior to use) and 0.45 μm (polyethersulfone membrane filter; Hyundai Micro). To minimize cell rupture, which could increase the measured AT and thereby result in biases in the AT values (Kim et al., 2006), the filtration was completed within 5 min under a weak vacuum pressure (<10 kPa) using an aspirator. This established method for filtration resulted in a negligible release of cellular dimethyl-sulfoniopropionate, which is a sensitive indicator of cell rupture (Park et al., 2014). To avoid pressurization and consequent cell rupture as a result of excessive accumulation of phytoplankton cells on the filter (Collos et al., 2014), we limited the filtration of seawater to a volume of 150 mL, which was sufficient for a single measurement for AT. The filtration was terminated when 10 mL of sample remained in the filter funnel.

2.3 Determinations of AT, AT−BIO, AT−PLANKTON, and AT−BACTERIA

For each seawater sample the AT values of an unfiltered and a filtered (0.7 and 0.45 μm) sample were measured within 12 h of collection. The total contribution of all biogenic particles to the AT of the unfiltered seawater (AT−BIO; Equation 2) was determined as the difference between the measured AT values of the unfiltered sample and the 0.45 μm-filtered sample, followed by subtraction of the CaCO3 contribution (AT−PIC), which was calculated from the particulate inorganic carbon (PIC) concentration determined as described in Text S1. In this calculation we assumed that the particulate matter contribution to AT mostly comprised phytoplankton and bacteria cells, and CaCO3 shells. The contribution of phytoplankton cells to AT (AT−PLANKTON; Equation 3) was determined as the difference in measured AT between the unfiltered sample and the 0.7 μm-filtered sample, followed by subtraction of AT−PIC. The contribution of bacteria (AT−BACTERIA; Equation 4) was determined by calculating the difference in measured AT values between the 0.7 μm-filtered sample and the 0.45 μm-filtered sample.

The precisions of measurements for three biological AT components were estimated from the root sum square of the standard deviations of all parameters involved (urn:x-wiley:00948276:media:grl63043:grl63043-math-0005; see those relevant parameters in Equations 2-4, and Text S2) and were found to be 1.1 μmol kg−1 for AT−BIO, AT−PLANKTON, and AT−BACTERIA. Note that the AT−PIC component was not considered in the estimation of the precisions for AT−BIO and AT−PLANKTON because measurable PIC values were found only for 6 of the 205 samples.

2.4 Choice of Carbonic Acid Dissociation Constants

As a reference set of constants for comparison, we chose the constants of Mehrbach et al. (1973) refitted by Dickson and Millero (1987) to different forms of functionality (referred to as MEHR73), because the MEHR73 is the set of constants that is consistent with laboratory (Lee et al., 1996; Lueker et al., 2000) and field measurements (Fong & Dickson, 2019; Lee et al., 2000; Millero et al., 2006). We compared CO2 parameter values calculated using MEHR73 with those calculated using the constants of Hansson (1973) refitted by Dickson and Millero (1987) (referred to as HAN73) and those calculated using the constants of Dickson and Millero (1987) (referred to as D&M87), which were the fits of the combined sets of data determined by Mehrbach et al. (1973) and Hansson (1973). The constants of Millero (2010) (referred to as M10), which were refits of data determined by Millero et al. (2006), were also used for comparison because M10 can be used over the most extensive range of salinity (1–50). The constants of Lueker et al. (2000) (referred to as LK00) were used. Note that the more recent sets of constants (M10 and LK00) agreed well with MEHR73.

3 Results

3.1 Contribution of Phytoplankton and Bacteria to AT

The AT−BIO values were significant in most surface ocean environments (Figure 1a). The median AT−BIO value across the study area was 2.4 μmol kg−1, but individual AT−BIO values varied widely, from <1.1 μmol kg−1 (close to the measurement precision of AT−BIO) to 18.8 μmol kg−1. The highest AT−BIO values found were for the North Pacific Ocean (median = 5.2 μmol kg−1), followed by the East Sea (median = 5.0 μmol kg−1), the Bering Sea (median = 2.6 μmol kg−1), the Chukchi Sea (median = 2.1 μmol kg−1), and the Arctic Ocean (median = 0.8 μmol kg−1). For 6 of the 205 samples, we found negative AT−BIO values that fell outside the AT measurement precision of −1.1 μmol kg−1. Such a small number of negative values indicates that minimal cell rupture occurred during the filtration of the seawater samples, and hence that cell rupture had a negligible effect on the AT−BIO values. Nonetheless, the negative AT−BIO values we found were probably more a result of the heterogeneity of the PIC particles in the seawater samples.

The AT−PLANKTON and AT−BACTERIA values were comparable, although local variability was found for both components. For AT−PLANKTON the highest levels were found for the East Sea (median = 3.7 μmol kg−1), followed by the North Pacific Ocean (median = 2.9 μmol kg−1), the Chukchi Sea (median = 1.3 μmol kg−1), the Bering Sea (median = 1.2 μmol kg−1), and the Arctic Ocean (median = 0.4 μmol kg−1). Similarly, for AT−BACTERIA the highest levels were found for the East Sea (median = 2.2 μmol kg−1) and the North Pacific Ocean (median = 2.2 μmol kg−1), followed by the Bering Sea (median = 1.1 μmol kg−1), the Chukchi Sea (median = 0.8 μmol kg−1), and the Arctic Ocean (median = 0.1 μmol kg−1; Figure 1b). In most basins investigated the AT−PLANKTON values were slightly higher than those of AT−BACTERIA. The AT−PIC was undetectable at most sampling locations except for a few where several micromolar levels of AT−PIC were detected.

For two Arctic Ocean locations we measured the AT−BIO down to 1,600 m depth (inset in Figure 1a). We found that it was insignificant except for waters shallower than 200 m depth, where organic particles were likely to be abundant. In waters deeper than 200 m, the AT−BIO values were near zero because of the microbial degradation of particulate organic matter.

3.2 AT−PLANKTON Versus Particulate Organic Carbon (POC)

In all five basins the AT−PLANKTON values for the same POC concentration varied by a factor of four or more (inset in Figure 2). The highest ratio of AT−PLANKTON to POC (μmol kg−1:μmol kg−1) was found for the East Sea (median = 0.19), followed by the North Pacific Ocean (median = 0.11). Three other basins (the Chukchi and Bering seas, and the Arctic Ocean) showed the lowest median values (0.07, 0.05, and 0.03, respectively), which for comparison are similar to the ratio of P. minimum (0.04) found in the monoculture experiments (Kim et al., 2006).

Details are in the caption following the image

The ratio of AT−PLANKTON to particulate organic carbon (POC) for all surface seawater samples for the five basins studied. The middle, lower, and upper horizontal lines in the boxes indicate the median, and the first and third quartiles of the ratio, respectively. The lower and upper whiskers show the minimum and maximum values, and the colored dots show the outliers. The thick horizontal gray lines indicate the ratio of AT−PLANKTON to POC for cultured Prorocentrum minimum (P. minimum; light gray) and Prorocentrum micans (P. micans; dark gray), reported by Kim et al. (2006). The inset shows AT−PLANKTON values as a function of POC concentrations; the slopes for P. minimum (light gray) and P. micans (dark gray) were obtained from Kim et al. (2006).

The AT−PLANKTON values were not significantly correlated with the POC concentrations, regardless of the basin investigated. For the East Sea the AT−PLANKTON value was highest (3.7 μmol kg−1), but the basin-median POC concentration (14.3 μmol kg−1) was the second lowest. For the Bering Sea the AT−PLANKTON value (1.2 μmol kg−1) was the second lowest, but the basin-median POC concentration (22.5 μmol kg−1) was the second highest.

4 Discussion

4.1 AT Variations Associated With Photosynthesis and Remineralization

The reactions of photosynthesis and respiration in the ocean, based on Redfield (1963), are represented by Equation 5.
During photosynthesis and respiration, change in the concentrations of nutrient species collectively contributes to seawater AT, as shown in Equation 6 (Wolf-Gladrow et al., 2007).
where TPO4 = [H3PO4] + [H2PO4] + [HPO42−] + [PO43−]; TNH3 = [NH3] + [NH4+]; and THNO2 = [NO2] + [HNO2]. The photosynthetic process increases AT by 1 mole for each mole of THNO2 or TPO4 assimilated, but decreases AT by 1 mole if TNH3 is assimilated as the nitrogen source.

Phytoplankton also tend to release organic acids during photosynthesis, some of which may subsequently dissociate into protons and conjugate bases. Unlike nutrient assimilation, these organic acids change neither AT nor electro-neutrality, but the protons that dissociate from them are likely to react with carbonate and borate ions, thus decreasing AC and AB but increasing AT−DOM; this was experimentally confirmed (Ko et al., 2016). This mechanism may also apply to phytoplankton and bacteria cells, but the exact cellular mechanism contributing to AT has not been elucidated.

The measurements of AT−BIO from the surface to 1,600 m depth (inset in Figure 1a) provide tantalizing hints regarding the fate of AT−BIO associated with the production and remineralization of organic matter. Higher values of AT−BIO in the surface tended to be associated with a higher abundance of phytoplankton and bacteria cells; however, the absence of AT−BIO in deep water constitutes a compelling evidence for the destruction of AT−BIO during organic matter remineralization in the ocean interior. Albeit derived from limited measurements, the depth-profile of AT−BIO may indicate its non-conservative nature in the ocean. More elaborate studies are needed to clarify these issues.

4.2 Distribution of AT−BIO in Coastal to Open Ocean Environments

The distribution of AT−BIO has been poorly documented in the ocean, and the results from two earlier studies are seemingly contradictory. In the productive bay study the AT−BIO ranged from 3 to 8 μmol kg−1 (Kim et al., 2006). In contrast, the study carried out in the oligotrophic North Pacific Ocean yielded insignificant AT−BIO values for both surface and subsurface waters (Chanson & Millero, 2007). Our AT−BIO values for productive coastal (adjacent to Korea and Japan) and open ocean waters (the North Pacific Ocean) are more consistent with the values found in the productive bay of Korea (Kim et al., 2006), while the low AT−BIO values we found for the other surface regions and deep waters are more similar to the results of Chanson and Millero (2007) for oligotrophic waters.

As the AT−PLANKTON values were derived mostly from phytoplankton cells, we anticipated a strong association between AT−PLANKTON values and POC concentrations. The absence of such an association is counterintuitive, but at the same time provides evidence that AT−PLANKTON is phytoplankton species-dependent. Rather than the absolute biomass of phytoplankton, differences in the characteristic surface area or shape of phytoplankton species could be more influential factors in determining the magnitude of AT−PLANKTON. Differing negative charge densities on cell surfaces among phytoplankton species could also explain the absence of the association, as some phytoplankton species are likely to have a higher negative charge density than others; in this case, more acids would be consumed during seawater titration, thereby resulting in higher AT−PLANKTON values. Explicitly, carboxyl groups having pK values of 4.5–6.0 can contribute more to AT as they are largely deprotonated at typical seawater pH, and thereby have more negative charge than other groups (Fong & Dickson, 2019; Ko et al., 2016). As the chemical composition of the cell surface is likely to differ with phytoplankton species, the differing ratios of AT to POC found for the study basins were likely to have arisen from differences in the dominant phytoplankton species present.

Based on these three studies carried out in a range of ocean environments, prediction of the distribution of AT−BIO in the ocean is difficult, because AT−BIO was found to be poorly correlated with other more easily measured parameters including POC and the chlorophyll-a concentration. The direct association between AT−PLANKTON and POC is ambiguous, as noted earlier. Prediction of AT−BACTERIA is even more difficult, because variations in this parameter have no association with either POC or AT−PLANKTON. Determination of the population of bacteria and the total bacterial surface area (Harris & Theriot, 2018; Takahashi et al., 2017) may provide the basis for estimating AT−BACTERIA, but this is not easily done concurrently with measurement of AT−BACTERIA in the field.

4.3 Significance of AT−BIO in Predictions of Carbon Parameters Using Thermodynamic Models Describing the Seawater Carbonate System

During our trans-Pacific Ocean study the AT−BIO measurements were not made concurrently with measurements of other CO2 parameters. Therefore, we could not directly evaluate the accuracy of the thermodynamic models used for describing the seawater CO2 system. As an alternative, for each basin we determined the differences in CT, calculated using the constants of MEHR73 and a basin median pH value and all our AT measurements in that basin, with AT−BIO included and excluded (Figure 3). We then determined the differences in calculated CT using the median values of pH and AT for the entire cruise track and five different sets of dissociation constants for carbonic acid (MEHR73, HAN73, D&M87, LK00, and M10). In the former calculation, the calculated CT values for both the East Sea and the North Pacific Ocean differed by as much as 10–17 μmol kg−1 between the case including AT−BIO and that excluding AT−BIO. The median differences in CT for the East Sea and the North Pacific Ocean basins were more equivalent to the differences found when the constants of HAN73 (and D&M87) versus MEHR73 were used. The median differences in CT for the Bering Sea, the Chukchi Sea and the Arctic Ocean were close to the differences found when the constants of M10 (and LK00) versus MEHR73 were used. With the exception of the East Sea and the North Pacific Ocean basins, the basin-median AT−BIO values were close to the estimated precision of AT−BIO (1.1 μmol kg−1). Nonetheless, the significant errors in the thermodynamic calculations based on AT measured using unfiltered surface samples need to be considered.

Details are in the caption following the image

Box plots representing differences in CT values calculated with AT−BIO included and excluded, based on a basin median pH value (8.1 for the East Sea, the North Pacific Ocean, the Bering Sea, and the Arctic Ocean; 8.3 for the Chukchi Sea) and all AT measurements for each basin using the constants of Mehrbach et al. (1973) as refit by Dickson and Millero (1987) (referred to as MEHR73). The middle, lower, and upper horizontal lines in the boxes indicate the median, and the first and third quartiles of the difference in CT, respectively. The lower and upper whiskers show the minimum and maximum values, and the colored dots show the outliers. The thick gray horizontal lines indicate the differences between the CT values calculated from the median values of pH (8.1) and AT (2168.0 μmol kg−1) for the entire cruise track using the constant of MEHR73 versus the constants of HAN73, D&M87, LK00, and M10. Note that the basin median pH value was calculated using data from GLODAP collected near the cruise track in each basin (Key et al., 2015; Olsen et al., 2016).

5 Conclusion

The present study in conjunction with two other studies in the bay of Korea (Kim et al., 2006) and an open ocean environment (Chanson & Millero, 2007), shows that the AT−BIO values (mostly by phytoplankton and bacteria) were highly variable (near zero to as high as 19 μmol kg−1) across diverse ocean environments ranging from coastal waters to varied open ocean habitats (e.g., subtropics, subarctic, and the Arctic Ocean). The AT−BIO values we found were significant but substantially smaller than those for AC and AB. We recommend that AT be measured using filtered seawater samples, particularly in productive ocean regions, because the influence of AT−BIO on the calculated parameters is considerably greater than the measurement precision. We note that discretion is needed when deciding whether to perform seawater filtration in less productive ocean environments, where seawater filtration may not greatly benefit.


We wish to gratefully acknowledge the constructive reviews of this paper by Andrew Dickson and an anonymous reviewer. This research was supported by National Research Foundation of Korea funded by the Ministry of Science and ICT (NRF-2021R1A2C3008748) and by the National Institute of Fisheries Science (R2021051). We greatly appreciate that the chief scientist and crew members of the R/V ARAON help seawater sampling. We also thank the NASA Goddard Space Flight Center, Ocean Ecology Laboratory, Ocean Biology Processing Group and the Global Ocean Data Analysis Project for making chlorophyll-a and pH data available.

    Data Availability Statement

    The chlorophyll-a data were obtained from the Modis-Aqua data set (https://oceancolor.gsfc.nasa.gov/data/10.5067/AQUA/MODIS/L3M/CHL/2018/). The pH data were downloaded from the GLODAPv2.2020 data set (https://www.glodap.info/index.php/merged-and-adjusted-data-product/). Measurements data for all seawater samples during the field study are available at the Dryad Digital Repository (https://doi.org/10.5061/dryad.fbg79cnvp).