Open Access
Add to BibSonomy
Abstract
Using biogeochemical-Argo float measurements, we propose, for the first time, an optical proxy for particulate organic nitrogen concentration (PON) in the Western Tropical South Pacific, an area influenced by dinitrogen (N2) fixation. Our results show a significant relationship between the backscattering coefficient at 700 nm (bbp) and PON, especially when the latter is measured using the wet oxidation method (R2=0.87). bbp may be used to estimate PON concentrations (PONopt) between 0.02 and 0.95 µM, allowing for unprecedented monitoring using autonomous profiling floats. The bbp vs PON relationship can be used to study phytoplanktonic biomass dynamics at relevant seasonal temporal scales, with clear evidence of PONopt as a proxy of phytoplanktonic biomass, at least for this specific area. Temporal analyses of PONopt show significant increases (from 0.16 to 0.80 µM) likely related to new production associated to N2 fixation events measured during stratification periods in the Melanesian Archipelago.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
The surface waters of the subtropical South Pacific gyre (SPG) are permanently depleted in dissolved inorganic nitrogen (NO3-) [1], making this area the largest oceanic desert in the world’s ocean [2]. However, in some oligotrophic areas, biological N2 fixation offers the marine phytoplankton community a mechanism to relieve nitrogen limitation in the euphotic surface layer [3,4]. In the framework of the Oligotrophy to UlTra-oligotrophy PACific Experiment (OUTPACE) cruise (19 Feb-3 Apr 2015), record N2 fixation rates were recently observed in the upper waters of the Western Tropical South Pacific (WTSP) at the end of austral summer [5]. Despite a NO3- depleted mixed layer, a significant increase of phytoplanktonic biomass was observed during N2 fixation events. However, the limited observations of seasonal phytoplankton biomass dynamics, hypothesized to be largely driven by N2 fixation [6,7], considerably restricts our understanding of the WTSP’s biogeochemical functioning. To overcome present limitations, in situ observations over a broad range of time scales are required.
Biogeochemical-Argo (BGC-Argo) profiling floats are capable of autonomously observing bio-optical properties such as Chlorophyll-a (Chla) fluorescence and particulate backscattering coefficient (bbp) at high frequency [8]. The use of bio-optical proxies has been previously shown to have high reliability in the estimation of biogeochemical variables such as Chla and particulate organic carbon (POC), among others [8,9]. While Chla fluorescence is the most commonly used proxy for living phytoplankton cells [10,11], bbp variability is instead driven by both the algal and non-pigmented particle pools including viruses, heterotrophic bacteria, and non-living cells in case 1 oceanic waters [12]. For this reason, bbp has long been used as a proxy of POC in open ocean water in the absence of mineral particles [9,12–19].
In oligotrophic areas, the carbon hydrogen nitrogen (CHN) method requires a high volume of filtered seawater, up to 10 L, to obtain accurate POC measurements [19]. Filtering such volumes of sea water is extremely time consuming and limits throughput of large numbers of samples. Several factors can lead to large biases in estimated POC in oligotrophic areas, including contamination, adsorption of dissolved organic carbon (DOC) onto filters, particle formation in bottle samples after collection, the contribution of particulate inorganic carbon, as well as particle retention, among others [19–22]. Particulate organic nitrogen and phosphorus measurements obtained by the wet oxidation method (PON(wet) and POP(wet), respectively) could represent a valuable alternative in oligotrophic areas, since the method’s sensitivity requires smaller volumes of seawater than the CHN method [23], making it less sensitive to the potential contaminations alluded to above. Wet oxidation requires a maximum volume of 1.2 L, even in oligotrophic areas. The smaller volume and the absence of an acidification step to remove inorganic particles on the filter significantly decrease both the time required for each sample and potential particle formation as transparent exopolymer particles (TEP; 22) in bottle samples, thus decreasing potential contamination.
The wet oxidation method has been shown to be more accurate than the CHN method for PON and POC measurements [23]. Based on the fact that POC and PON generally covary [1,6], one may expect a good correlation between bbp and PON. As a result, the use of an optical proxy of PON (PONopt) could provide a means to estimate seasonal variations of phytoplanktonic biomass and associated particles, especially in oligotrophic areas.
In areas where nitrogen is the key limiting nutrient, and N2 fixation is sufficiently favored, both an increase in PON standing stocks and a tight coupling between N2 fixation and surface water PON accumulation have been observed [24,25]. While PON estimates allow one to follow phytoplanktonic biomass, observations in the South Pacific are sparse [26] [Fig. 1(a)], stressing the need to develop indirect methods for estimating this key variable at regional and pertinent time scales. For this purpose, three BGC-Argo floats were deployed during the OUTPACE cruise. Our study focuses on the mixed layer, where the N2 fixation process mainly occurs [25]. The main goals of this work are (1) to characterize and discuss, for the first time, bbp vs PON relationships to define an optical proxy of PON (PONopt), and then, (2) to investigate the seasonal dynamics of the particulate organic biomass in the WTSP, an area influenced by N2 fixation events, using Chla and PONopt seasonal distributions.
Fig. 1. (a) Trajectory of the BGC-Argo floats deployed in this study (green triangle = FA, red triangle = FB, blue triangle = FC). The location of the OUTPACE section is represented by black (Western and Eastern Melanesian Archipelago) and grey (South Pacific gyre) circles. Open circles represent the location of available in situ PON measurements in the South Pacific [24]. (b) Zoom on the trajectory of the BGC-Argo floats. Colors represent the time (years) and the black dashed boxes show the “bloom periods” defined in the text.
2. Materials and methods
The OUTPACE cruise took place along a West to East transect [Fig. 1(b)]. A total of 18 stations were sampled from the oligotrophic water of the Western and Eastern Melanesian Archipelago (WMA and EMA, respectively) to the clearest ocean waters of the SPG [27].
PON and particulate organic phosphorus (POP) samples were collected at 16 depths between the surface and 500 m from a SBE 911+ CTD-Rosette in polycarbonate bottles. PON and POP concentrations were quantified spectrophotometrically following the wet oxidation method (PON(wet) and POP(wet)) based on persulfate digestion at 120 °C [28]. Following this method, a volume of 1.2 L was filtered through a pre-combusted (24 h, 450 °C) 47 mm GF/F filter. The filter was then placed in a Teflon bottle in which 20 mL of milli-Q water and 2.5 mL of the oxidizing reagent were previously dispensed (concentration factor: 1.2/0.0225). Nitrate and phosphate concentrations were then determined in the digested sample using an automated colorimetric procedure on a Technicon auto-analyzer [29]. The repeatability, calculated as the coefficient of variation (CV) for PON(wet) and POP(wet) field-collected replicates (n=10), was 2% and 3%, respectively. The accuracy, linked to the uncertainty of the calibration curve’s slope (calculated as the CV of the slope) of PON(wet) and POP(wet) was 1.46% and 1.36% (n=14), respectively. For each station, pre-combusted GF/F filters were used on board as sample blanks. The blank consisted of adding the same volume of oxidizing reagent to 20 mL of milli-Q water, in which a pre-combusted GF/F filter was previously added. The means of filter blanks of PON(wet) and POP(wet) were 0.036 ± 0.002 µM and 0.0021 ± 0.0001 µM. The quantification limits of PON(wet) and POP(wet), calculated as ten times the standard deviation of 10 blank measurements [30], were 0.02 µM and 0.001 µM, respectively. The maximum quantification limits for PON(wet) and POP(wet) were [PON]max = [standard NO3-]max/(concentration factor) = 0.95 µM and [POP]max = [standard PO43-]max/(concentration factor) = 0.060 µM, respectively. Mineralization efficiencies measured daily with P-choline and urea standards were 100 ± 2% for N and 99 ± 1% for P (n=27).
A second set of PON measurements used a PerkinElmer 2400 CHN analyzer (PON(CHN)), with standards prepared with a 20 g L-1 glycine (VWR C14037000) solution (N range: 0.15-10 µM). Seawater samples (2 L) were filtered through pre-combusted (4 h, 450° C) 25 mm GF/F filters, dried at 60° C and stored in 1.5 mL Eppendorf PE tubes. The repeatability for PON(CHN) field-collected replicates (n=6) was 4.60% and the calibration curve slope CV was 2.83% (n=13). For each station, pre-combusted GF/F filters were used on board as sample blanks. These filter blanks were processed in the same way as sample filters without the filtration step. The mean of filter blanks of PON(CHN) was 0.021 ± 0.014 µM. The quantification limit of PON(CHN) was 0.13 µM. PON(wet) and PON(CHN) concentrations showed an excellent agreement (R2=0.92; slope = 1.03 ± 0.03). The recovery between PON(CHN) and PON(wet) is close to 100% (95% C.I of the slope = [0.96 1.11]; 95% C.I of the intercept = [-0.03 0.028]) with a very low standard deviation over the range of concentrations measured as part of this study, highlighting that the wet oxidation method recovered all the PON from living phytoplankton and associated particles as previously reported by [23,31].
Three BGC-Argo floats (FA), (FB) and (FC) were deployed in March/April 2015 near the stations LDA, LDB and LDC respectively [Fig. 1(b)], and their collected data were downloaded from the Coriolis database website (ftp://ftp.ifremer.fr/ifremer/argo/dac/coriolis/). These floats were equipped with a Sea-Bird Electronics (SBE41CP) conductivity-temperature-depth (CTD) sensor (Seabird Inc., USA) and an additional sensor package: the WETLabs Environmental Characterization Optics triplet puck (ECO3, Seabird Inc., USA) measuring the fluorescence of Chla at excitation/emission wavelengths of 470/695 nm and the angular backward scattering coefficient of particles at 700 nm (Table 1).
Table 1. Equipment details for each float used in this study, (the first fifteen, nineteen, and eleven profiles were recorded every day for floats FA, FB and FC, respectively, before starting to sample every five days).
Measurements were collected every 1 or 5 days between 1,000 dbar and the surface, with a sampling resolution of 1 dbar between the surface (0.1-1 dbar) and 250 dbar and 10 dbar between 250 and 500 dbar, respectively. Growth rate of diazotrophs like Trichodesmium (the main diazotroph in the WTSP) are generally low and blooms last for months, indicating that 5 days is a correct sampling time to capture a diazotroph bloom. The data were quality controlled following the standard Argo protocol [32–34].
Mixed layer depth (MLD) was calculated using a threshold density of 0.03 kg m-3 deviation from the reference value at 10 m depth [35].
The fluorescence measurements of Chla were converted to Chla concentrations according to the procedure detailed by [33]. Chla concentrations were cleaned from out-of-range values and, following the recommendations of [36], adjusted Chla concentrations were divided by a factor of two. A non-photochemical quenching correction was applied following the standard Argo protocol [33]. Vertical profiles of Chla concentrations showed that most of the Chla values observed in deep waters (> 200 dbar) were negative (figure not shown). To correct the negative deep Chla concentrations, we removed a constant value (the deepest Chla fluorescence value from the profile, i.e., so-called “deep-offset correction”). During the OUTPACE cruise, a “Fluorimeter, Chelsea Aquatracka MKIII” attached to a SeaBird CTD rosette was used to measure the Chla fluorescence. Calibration of the fluorimeter was carried out using HPLC Chla measurements from 13 OUTPACE stations (stations SD1, LDA, SD6, SD8, SD10, SD12, LDB, LDC, SD14, SD15), hence just before the BGC-Argo deployments. Very good agreement (p>0.01, Fig. 2) was observed between the first vertical profile of adjusted Chla concentrations and the vertical profile of Chla concentrations measured in situ at the closest OUTPACE CTDs in time and space (Table 1).
Fig. 2. (a, b, c) Comparison between the respective concentrations of Chla (mg m-3) measured by the CTD 067/151/199 (black dashed lines), uncorrected Chla (mg m-3) concentrations measured by floats FA / FB / FC (gray lines), and corrected (colored lines) Chla, depending on the density (kg m-3).
The backscattering sensors measure the angular scattering coefficient at 124° relative to the direction of light propagation at a wavelength of 700 nm. This measurement is then transformed into the bbp following [34], using the conversion factor of [37] (χ = 1.076). Negative values of bbp were removed and vertical profiles were quality-controlled following the standard Argo protocol [34]. Both datasets (Chla and bbp) have indeed been checked qualitatively and no sensor drift or bio-fouling have been observed.
The variability of bbp with PON is investigated between the surface and 500 dbar using the OUTPACE CTD-rosette water samples, which were the closest in time and space to the float profiles where bbp was measured (Table 1). To avoid the effect of internal waves, each bottle data value was paired with a bbp value at the same density coordinate. Differences in density between the bottle data and associated bbp values were found to be less than 0.005 kg m-3.
3. Results
3.1 From backscattering to particulate organic matter
The bbp vs PON relationship is investigated to understand the extent to which bbp can be used as a proxy of PON to better assess the impact of N2 fixation events at the relevant seasonal temporal scales. Here we investigate the bbp vs PON relationship (1) in the mixed layer (ML), (2) in the ML and the deep Chla maximum (DCM) [0-150 dbar, Fig. 3(b)], and (3) in the whole sampled water column (0-500 dbar), before deriving PONopt and POPopt.
Fig. 3. (a) Ratio of bbp/PON(wet) vs Pressure, (b) Chla concentrations vs Pressure, and scatter plots between bbp (m-1) measured during the first profiles of the FA, FB and FC floats and (c) PON(wet), (d) PON(CHN), and (e) POP(wet). (The yellow, grey, and black lines represent the best linear regression fits for the data points in the ML, 0-150 dbar, and 0-500 dbar, respectively). Mixed layer = yellow markers/yellow lines, 0-150 dbar = yellow + grey markers/ grey lines, 0-500 dbar = yellow + grey + black markers/ black lines. (f) bbp vs PON(wet) slopes in the mixed layer, between the surface and 150 dbar, and between the surface and 500 dbar (note: circle = FA/LDA, square = FB/LDB, triangle = FC/LDC). Red lines correspond to quantification limits. Statistical parameters are in Table 2.
Table 2. Statistical parameters for bbpvs PON(wet), PON(CHN), and POP(wet) for the three depth categories: 0-500 dbar, 0-150 dbar, and Mixed Layer. n is the number of data points, R2 is the determination coefficient, RMSE the Roots Mean Square Error (µM), Median of residuals (residuals = observed minus fitted values), and 75th and 25th percentile represents the interquartile range of residuals).
All PON(wet) concentrations and most of the PON(CHN) concentrations are above their respective quantification limits. By considering these data, a significant relationship is obtained between bbp and PON, regardless of the experimental method used to measure PON (wet vs CHN) (Fig. 3; Table 2). The bbp vs PON(wet) relationship is better than that of bbp vs PON(CHN) between the surface and 150 (500) dbar [Figs. 3(c), 3(d); Table 2]. Methodologically based variability could explain this discrepancy. Indeed, the wet oxidation method achieves a sensitive measurement from a smaller volume (1.2 L) of seawater than the CHN method (2-10 L), making this method highly sensitive and suitable for PON analyses in oligotrophic waters [6,28].
The bbp/PON(wet) ratios exhibit a low variability with depth, with an average value of 1.9 × 10−3 ± 5.7 × 10−4 m-1 µM-1 [Fig. 3(a)] in the layer including the ML and the DCM, [Fig. 3(b)]. In the ML, the slope derived from the bbp vs PON(wet) relationship [1285 ± 114; Figs. 3(c), 3(f)] is not significantly different from that observed in the 0-150 dbar layer [1280 ± 98; Fisg. 3(c), 3(f)]. In contrast, below the DCM (>150 dbar), an increase in bbp/PON(wet) ratios is observed, with an average value reaching 4.1 × 10−3 ± 9.5 × 10−4 m-1 µM-1 [Fig. 3(a)]. The slope from the bbp vs PON(wet) relationship obtained between 0 and 500 dbar is significantly lower than those observed in both the ML and between the surface and 150 dbar [Fig. 3(f)].
Hereafter, to investigate the temporal variability of phytoplanktonic biomass and associated particles in the WTSP, the bbp measurements will be discussed in terms of PON using the relationship established from the bbp and PON(wet) data set measured between the surface and 150 dbar [Fig. 3(c)] [Eq. (1)].
Studying each float separately in the 0-150 m layer, the bbp vs PON(wet) relationships are only significant for FA and FB, with similar slopes (p<0.001) (Table 3). The FC/LDC data are very few (n=9) and FC exhibits low and quasi-constant PON(wet) values. We therefore decide not to calculate the bbp vs PON relationship for the LDC station individually. Nevertheless, including FC/LDC data in the global bbp vs PON(wet) relationship does not significantly affect the slope (Table 3), so the global relationship is therefore calculated with all the data in the 0-150 m layer.Table 3. Slopes and intercepts for the bbpvs PON(wet) and bbpvs POP(wet) relationships for FA-FB/LDA-LDB, and FA/FB/FC in the 0-150 m layer. (R2 = the determination coefficient, p is the p-value).
We consider the intercept of the PON(wet) vs bbp relationship as the lower detection limit of bbp to derive PONopt. The highest value of bbp for which this equation is valid corresponds to the maximum value measured. Values of bbp for which the bbp vs PON(wet) relationship is valid range between 3.4×10−4 and 9.5×10−4 m-1. Additionally, values of bbp for which the bbp vs POP(wet) relationship is valid range between 2.7×10−4 and 9.5×10−4 m-1.
3.2 Mixed layer depth and chlorophyll-a seasonal distribution
In general, the seasonal Chla patterns mimic the MLD seasonal variability, with low mixed layer values observed during the summer (January/February/March) and high values during the winter (July/August) (Fig. 4). However, in the WMA (FA), increases of Chla at shorter time scales were superimposed onto the seasonal pattern, with values reaching 0.09 mg m-3 in March/April 2015 and 0.10 mg m-3 in October/November 2015. These episodes were not related to abrupt changes in the MLD but occurred during conditions of shallow MLD (< 60 dbar) [dashed frame on Fig. 4(a)]. In the EMA (FB), Chla followed the same seasonal variations as observed in 2016 and 2017 in the WMA (FA), except for the summer period: despite a shallow MLD (< 45 dbar), Chla showed relatively high values, reaching 0.14 mg m-3 in March/April 2015, 0.08 mg m-3 in January/February 2016, and 0.10 mg m-3 in February 2017 [dashed frame on Fig. 4(c)]. The absence of intense vertical mixing and relatively high light levels (figure not shown) during these stratified periods should result in a diminution of the Chla concentrations. Observing the opposite, we hypothesize that the high Chla concentrations point to an additional controlling factor of the variability of Chla dynamics in the area, probably linked to diazotroph blooms.
Fig. 4. Temporal variations of the mixed layer depth (dbar) (black line) and the mixed layer average Chla (mg m-3) concentration for floats (a) FA (green line), (c) FB (red line), and (e) FC (blue line), along with temporal variations of (b) the mixed layer average PONopt concentration (µM) (± sd) for float FA (green line), (d) float FB (red line), and (f) float FC (blue line). Lighter colors show standard deviation around the mean (the uncertainty in the regression has been incorporated into the error propagation). Black dashed frames show the “bloom periods” as defined in the text.
3.3 PON seasonal distribution
In contrast to Chla, the temporal evolution of PONopt was relatively even, with low values generally below 0.20 µM (Fig. 4). However, as for Chla, increases of PONopt were observed during the first year in the WMA (FA) and each year at particular times, between January and April, in the EMA (FB). Indeed, in the WMA (FA), between January 2016 and July 2017, despite the seasonal cycle of Chla, PONopt was constant over time [Fig. 4(b)], with an average value of 0.18 ± 0.06 µM over this period [Fig. 4(b)]. Conversely, PONopt values were significantly higher (p<0.01, t-test) in 2015 than those observed in 2016 and 2017 [Fig. 4(b)]. Maximum values were recorded during March/April 2015 and October/November 2015, averaging 0.41 ± 0.09 µM during both periods [dashed frames on Fig. 4(b)], simultaneous with increases of Chla [dashed frames on Fig. 4(a)].
In the EMA (FB) during the summer periods, the average PONopt values reached 0.70 ± 0.10 µM during March/April 2015, 0.50 ± 0.20 µM during January/February 2016 and 0.60 ± 0.03 µM during February 2017 [dashed frames on Fig. 4(d)], coinciding with relatively strong increases of Chla [dashed frames on Fig. 4(c)]. Outside of these summer periods, PONopt values were mostly constant with an average value of 0.17 ± 0.06 µM [Fig. 4(d)], similar to FA in 2016/2017. In the SPG (FC), no increases of PONopt were observed during the summer period. Instead, PONopt was constant throughout the study period, with an average value of 0.16 ± 0.04 µM [Fig. 4(f)].
Hereafter, periods with both significant increases of PONopt by a factor of 5 (from 0.16 to 0.80 µM) and Chla by a factor of 4 to 5 (from 0.03 to 0.14 mg m-3) are called ‘bloom periods’ [dashed frames on Fig. 1(b) and Fig. 4]. These periods are: March/April 2015 and October/November 2015 for FA, and March/April 2015 and January/February 2016 and February 2017 for FB.
4. Discussion
At first order, bbp variability is driven by the concentration of bulk particulate matter, with changes in the particulate size distribution, refractive index, and particle morphology, among other factors, generally acting at second order [12]. The inherent optical properties (IOP) values relative to the concentration (or mass) of biogeochemical parameters (here bbp/PON(wet)) are sensitive to these second-order effects. While a relatively constant bbp/PON(wet) ratio indicates homogeneous bulk particulate matter in terms of backscattering efficiency in the surface layer, an increase at depth indicates modification and/or heterogeneity of bulk particulate matter. Our result suggests that the influence of different particle composition and/or a potential presence of distinct microbial communities observed during the OUTPACE cruise (diazotrophic organisms in the ML and DCM microbial communities below the ML, [25,38]) was not significant compared to changes in organic particulate concentration. Thus, it seems that biomass changes are a dominant source of bbp variability, as previously reported by [39]. These observations are also in agreement with those of [19], who reported that the POC vs bbp relationship was relatively insensitive to community composition. In contrast, below the DCM (>150 dbar), an increase in bbp/PON(wet) ratios is observed. This finding could reflect the presence of an assemblage of particles dominated by detritus and heterotrophs (but requires further analysis based on a better characterization of the bulk particulate matter below the DCM). The main conclusion of these observations is that the highly sensitive wet oxidation method provides a good bbp vs PON(wet) relationship for this area between the surface and 150 dbar, and valid between 0.02 and 0.95 µM (∼50-fold biomass increase), which thanks to float-based measurements of bbp can be leveraged to expand both the number of observations and the spatiotemporal scales resolved.
During the periods of no bloom in the Melanesian Archipelago, and throughout the study period in the SPG, PONopt concentrations were remarkably stable and low (<0.20 µM), displaying a net balance between production (or supply) and remineralization (or removal) processes [24]. The relative increase of Chla and the constant values of PONopt during the winter periods suggest a seasonal change in Chla cell quota due to a lower light intensity in winter coupled with deeper mixed layer depths [40,41]. As a result, the constant PONopt suggests that the marked Chla seasonal cycle, which can be falsely interpreted as a biomass variation, is in fact not related to new production but is instead a steady phytoplankton biomass. In our study region, the absence of phytoplanktonic blooms could be linked to nitrogen limitation [1], while N2 fixation could be constrained by lower iron availability [42,43].
During the bloom periods of the austral summer, coinciding with the increase of Chla concentrations, PONopt concentrations increased by a factor of two to five only in the Melanesian Archipelago. A co-variation between Chla concentrations and bbp values (or PONopt in our case) is linked to phytoplankton biomass production [44–47]. The results obtained in this study extend this observation to subtropical areas, pointing to outstanding increases of living phytoplanktonic biomass and associated particles, even in oligotrophic regions. To explain the recurrent increases of phytoplanktonic biomass during austral summer periods, it is necessary to identify the nitrogen sources in the mixed layer. The two first blooms (March/April 2015 for both FA and FB) reported in this study occurred during the OUTPACE cruise. During those blooms, high N2 fixation rates were reported while NO3- was extremely low. Regarding the diazotroph bloom observed at station LDB, mesoscale vertical fluxes were too weak to displace the nitracline [48]. As a result, delivered nitrogen to the surface from mesoscale activity could not sustain the observed primary production. Regarding the blooms observed in our study after the cruise, except during February 2017 in the EMA no significant decreases of salinity were measured in parallel to PONopt increases over the study period [Fig. 5], suggesting no direct influence of precipitation or riverine inputs.
Fig. 5. Temporal variations of (a) the average salinity (±sd) in the mixed layer for float FA (green line) and (b) float FB (red line). Black dashed frames show the “bloom periods” defined in the text.
Furthermore, atmospheric particle deposition fluxes are very low in the region [49]. The top of the nitracline was 90 dbar in the area during the summer [6], significantly deeper than the measured MLD (<60 dbar, Fig. 4). However, physical processes could vertically displace isopycnals tens of meters along with the nitracline [50]. Nevertheless, extreme events such as cyclones reported in this area perturbed the phosphacline but not the deeper nitracline, resulting in surface increases of phosphate but not of nitrate [51]. Therefore, in the absence of vertical N supply, the variations of PONopt appear to be a valuable proxy of new production related to intense N2 fixation events during austral summer conditions. In this specific environment characterized by active biological N2 fixation, most of the new nitrogen is rapidly integrated into the phytoplanktonic biomass. Even if part of the nitrogen is released to the labile dissolved organic and inorganic pools, it is rapidly re-assimilated by the nitrogen-starved organisms [52,53]. We therefore conclude that the increases of PONopt are most likely related to living POM, i.e. phytoplankton cells, and, to a lesser extent detritus.
In our study, we propose to use an optical proxy of PON rather than the traditional optical proxy of POC for several reasons. First, POC values show concentrations below 4 µM in oligotrophic gyres [54]. This limit value is close to the uncertainty reported in previous studies predicting POC from bbp [9,19]. The use of PON would be a superior alternative as the wet oxidation procedure, coupled with spectrophotometric measurements, allows accurate quantification of PON (quantification limit = 0.02 µM which correspond to 0.13 µM in terms of POC following the Redfield proportion), with reasonable volumes of filtered seawater (1.2 L). Second, oceanic primary production is primarily controlled by nitrogen availability [55,56]. Thus, it would be helpful, in order to better understand controls of primary production, to gain insight on nitrogen cycling, particularly on the seasonal cycle of nitrogen pools directly, rather than carbon pools. This argument is even more important to consider in environments characterized by active biological N2 fixation [5] where N budgets are needed [6]. Theoretically, all essential elements may be used to track primary production rates and biomass accumulations. Relative to carbon, the ∼30 essential nutrient elements display a range of plasticity in their cellular requirements [57], with N being the least plastic by varying in cellular N:C molar ratios from ∼1/5 to 1/10, a factor of two [58]. This variability is low compared to changes in biomass at the scale of the world Ocean, and even in the oligotrophic WTSP [6]. Therefore, if biomass is dominated by phytoplankton, a close link between PON and POC is to be expected. Similarities notwithstanding, the relative ease of PON measurements in oligotrophic areas favors the use of a PON optical proxy.
In addition to PON(wet), the wet oxidation method can simultaneously measure POP(wet). Interestingly, the bbp vs POP(wet) relationship was statistically better than bbp vs PON(wet) [Fig. 3; Table 2], highlighting the possible use of POP as a biomass proxy, which would drive an accurate estimation of standing POP stocks and associated biogeochemical fluxes. The better relationship between POP and bbp than PON and bbp suggests that the bulk particulate matter of POP could be less optically variable than PON-bulk particulate matter. This result could be linked to the variability of particle composition between both PON and POP pools. One main signature of a given particulate organic matter pool is the relative contributions of living particles and detrital material. Turnover rates of POP have been previously shown to be significantly higher than POC in the subtropical south Pacific [59], with the carbon and nitrogen pools containing more refractory material than the phosphorus pool [60–62]. Consequently, POP was considered to contain a higher contribution of living particles than PON or POC [59]. Therefore, the significant bbp vs POP(wet) relationship obtained [Fig. 3(c); Table 2] suggests that the variability of bbp could be linked, at least for this study area, more to the abundance of phytoplanktonic biomass than to associated material, as previously reported by [63] and [64]. Consequently, the optical proxy of POP estimated from bbp could also be a living biomass indicator and offer its own unique perspectives.
5. Conclusion
The combined use of in situ PON (and POP) measurements using the wet oxidation method, and optical properties measured by BGC-Argo floats close to in situ measurements, allowed to highlight, for the very first time, the excellent bbp vs PON (and POP) relationships between 0 and 150 dbar in under-sampled oligotrophic waters. In such areas, the quantification of PONopt, in the range between 0.02 and 0.95 µM, documents phytoplanktonic biomass dynamics (and associated properties) at the relevant seasonal temporal scale. Increases of PONopt by a factor of 5 (from 0.16 to 0.80 µM) were observed in the mixed layer of the WTSP during stratified conditions in the absence of significant nitrogen sources other than N2 fixation. The pertinence of the bbp vs PON (POP) relationships should be investigated in other oligotrophic areas, but also in other trophic regimes. This new relationship also opens a promising avenue to assess PON (POP) from ocean color remote sensing using various existing inverse methods [65–67].
Funding
Centre National de la Recherche Scientifique (ANR-14-CE01-0007-01).
Acknowledgments
The authors thank the crew of the RV L'Atalante for outstanding shipboard operations. Gilles Rougier and Marc Picheral are warmly thanked for their efficient help in CTD rosette management and data processing, as well as Catherine Schmechtig for the LEFE-CyBER database management. All data and metadata are available at the following web address: http://www.obs-vlfr.fr/proof/php/outpace/outpace.php. The Argo data were collected and made freely available by the International Argo Project and the national programmes that contribute to it (http://www.argo.ucsd.edu, http://argo.jcommops.org). Argo is a pilot programme of the Global Ocean Observing System. The authors would like to acknowledge Sandra Helias Nunige and Karine Leblanc for providing particulate organic concentrations.
This is a contribution of the OUTPACE (Oligotrophy from Ultra-oligoTrophy PACific Experiment) project (https://outpace.mio.univ-amu.fr/) funded by the French research national agency (ANR-14-CE01-0007-01), the LEFE-CyBER program (CNRS-INSU), the GOPS program (IRD), the CNES, and from the European FEDER Fund under project 1166-39417. The OUTPACE cruise (http://dx.doi.org/10.17600/15000900) was managed by the MIO (OSU Institut Pytheas, AMU, CNRS) from Marseilles (France)
Disclosures
The authors declare no conflicts of interest.
References
1. P. Raimbault, N. Garcia, and F. Cerutti, “Distribution of inorganic and organic nutrients in the South Pacific Ocean – evidence for long-term accumulation of organic matter in nitrogen-depleted waters,” Biogeosciences 5(2), 281–298 (2008). [CrossRef]
2. A. Morel, H. Claustre, and B. Gentili, “The most oligotrophic subtropical zones of the global ocean: Similarities and differences in terms of chlorophyll and yellow substance,” Biogeosciences 7(10), 3139–3151 (2010). [CrossRef]
3. R. C. Dugdale, D. W. Menzel, and J. H. Ryther, “Nitrogen fixation in the Sargasso Sea,” Deep Sea Res., Part I 7(4), 297–300 (1961). [CrossRef]
4. D. Karl, A. Michaels, B. Bergman, D. Capone, E. Carpenter, R. Letelier, F. Lipschultz, H. Paerl, D. Sigman, and L. Stal, “Dinitrogen fixation in the world’s oceans,” Biogeochemistry 57(1), 47–98 (2002). [CrossRef]
5. S. Bonnet, M. Caffin, H. Berthelot, and T. Moutin, “Hot spot of N2 fixation in the western tropical South Pacific pleads for a spatial decoupling between N2 fixation and denitrification,” Proc. Natl. Acad. Sci. U. S. A. 114(14), E2800–E2801 (2017). [CrossRef]
6. T. Moutin, T. Wagener, M. Caffin, A. Fumenia, A. Gimenez, M. Baklouti, P. Bouruet-Aubertot, M. Pujo-Pay, K. Leblanc, D. Lefevre, S. Helias Nunige, N. Leblond, O. Grosso, and A. de Verneil, “Nutrient availability and the ultimate control of the biological carbon pump in the Western Tropical South Pacific Ocean,” Biogeosciences 15(9), 2961–2989 (2018). [CrossRef]
7. A. Gimenez, M. Baklouti, T. Wagener, and T. Moutin, “Diazotrophy as the main driver of the oligotrophy gradient in the western tropical South Pacific Ocean: results from a one-dimensional biogeochemical–physical coupled model,” Biogeosciences 15(21), 6573–6589 (2018). [CrossRef]
8. D. Roemmich, M. H. Alford, H. Claustre, K. Johnson, B. King, J. Moum, and M. Scanderbeg, “On the Future of Argo: A Global, Full-Depth, Multi-Disciplinary Array,” Front. Mar. Sci. 6, 439 (2019). [CrossRef]
9. K. S. Johnson, J. N. Plant, L. J. Coletti, H. W. Jannasch, C. M. Sakamoto, S. C. Riser, D. D. Swift, N. L. Williams, E. Boss, N. Haentjens, L. D. Talley, and J. L. Sarmiento, “Biogeochemical sensor performance in the SOCCOM profiling float array,” J. Geophys. Res.: Oceans 122(8), 6416–6436 (2017). [CrossRef]
10. J. J. Cullen, “The Deep Chlorophyll Maximum: Comparing Vertical Profiles of Chlorophyll a,” Can. J. Fish. Aquat. Sci. 39(5), 791–803 (1982). [CrossRef]
11. D. A. Siegel, M. J. Behrenfeld, S. Maritorena, C. R. McClain, D. Antoine, S. W. Bayley, P. S. Bontempi, E. S. Boss, H. M. Dierssen, S. C. Doney, R. E. Eplee Jr, R. H. Evans, G. C. Feldman, A. Fields, B. A. Franz, N. A. Kuring, C. Mengelt, N. B. Nelson, F. S. Patt, W. D. Robinson, J. L. Sarmiento, C. M. Swan, P. J. Werdell, T. K. Westberry, J. G. Wilding, and J. A. Yoder, “Regional to global assessments of phytoplankton dynamics from the SeaWiFS mission,” Remote Sens. Environ. 135, 77–91 (2013). [CrossRef]
12. D. Stramski, E. Boss, D. Bogucki, and K. J. Voss, “The role of seawater constituents in light backscattering in the ocean,” Prog. Oceanogr. 61(1), 27–56 (2004). [CrossRef]
13. X. Xing, G. Qiu, E. Boss, and H. Wang, “Temporal and vertical variations of particulate and dissolved optical properties in the South China Sea,” J. Geophys. Res.: Oceans 124(6), 3779–3795 (2019). [CrossRef]
14. D. Stramski, “Refractive index of planktonic cells as a measure of cellular carbon and chlorophyll a content,” Deep Sea Res., Part I 46(2), 335–351 (1999). [CrossRef]
15. D. Stramski, R. A. Reynolds, M. Babin, S. Kaczmarek, M. R. Lewis, R. Rottgers, A. Sciandra, M. Stramska, M. S. Twardowski, B. A. Franz, and H. Claustre, “Relationships between the surface concentration of particulate organic carbon and optical properties in the eastern South Pacific and eastern Atlantic Oceans,” Biogeosciences 5(1), 171–201 (2008). [CrossRef]
16. H. Loisel, E. Bosc, D. Stramski, K. Oubelkheir, and P. Y. Deschamps, “Seasonal variability of the backscattering coefficient in the Mediterranean Sea based on Satellite SeaWiFS imagery,” Geophys. Res. Lett. 28(22), 4203–4206 (2001). [CrossRef]
17. H. Loisel, J. M. Nicolas, P. Y. Deschamps, and R. Frouin, “Seasonal and inter-annual variability of particulate organic matter in the global ocean,” Geophys. Res. Lett. 29(24), 49–49-4 (2002). [CrossRef]
18. L. Duforêt-Gaurier, H. Loisel, D. Dessailly, K. Nordkvist, and S. Alvain, “Estimates of particulate organic carbon over the euphotic depth from in situ measurements: Application to satellite data over the global ocean,” Deep Sea Res., Part I 57(3), 351–367 (2010). [CrossRef]
19. I. Cetinić, M. J. Perry, N. T. Briggs, E. Kallin, E. A. D’Asaro, and C. M. Lee, “Particulate organic carbon and inherent optical properties during 2008 North Atlantic Bloom Experiment,” J. Geophys. Res. 117(C6), C06028(2012). [CrossRef]
20. S. B. Moran, M. A. Charette, S. M. Pike, and C. A. Wicklund, “Differences in seawater particulate organic carbon concentration in samples collected using small-and large-volume methods: the importance of DOC adsorption to the filter blank,” Mar. Chem. 67(1-2), 33–42 (1999). [CrossRef]
21. W. D. Gardner, M. J. Richardson, C. A. Carlson, D. Hansell, and A. V. Mishonov, “Determining true particulate organic carbon: bottles, pumps and methodologies,” Deep Sea Res., Part II 50(3-4), 655–674 (2003). [CrossRef]
22. Z. Liu, G. Stewart, J. K. Cochran, C. Lee, R. A. Armstrong, D. J. Hirschberg, B. Gasser, and J. C. Miquel, “Why do POC concentrations measured using Niskin bottle collections sometimes differ from those using in-situ pumps?” Deep Sea Res., Part I 52(7), 1324–1344 (2005). [CrossRef]
23. P. Raimbault, F. Diaz, W. Pouvesle, and B. Boudjellal, “Simultaneous determination of particulate organic carbon, nitrogen and phosphorus collected on filters, using a semi-automatic wet-oxidation method,” Mar. Ecol.: Prog. Ser. 180, 289–295 (1999). [CrossRef]
24. D. V. Hebel and D. M. Karl, “Seasonal, interannual and decadal variations in particulate matter concentrations and composition in the subtropical North Pacific Ocean,” Deep Sea Res., Part II 48(8-9), 1669–1695 (2001). [CrossRef]
25. S. Bonnet, M. Caffin, H. Berthelot, O. Grosso, M. Benavides, S. Helias-Nunige, C. Guieu, M. Stenegren, and R. A. Foster, “In-depth characterization of diazotroph activity across the western tropical South Pacific hotspot of N2 fixation (OUTPACE cruise),” Biogeosciences 15(13), 4215–4232 (2018). [CrossRef]
26. A. C. Martiny, J. A. Vrugt, and M. W. Lomas, “Concentrations and ratios of particulate organic carbon, nitrogen, and phosphorus in the global ocean,” Sci. Data 1(1), 140048 (2014). [CrossRef]
27. T. Moutin, A. M. Doglioli, A. De Verneil, and S. Bonnet, “The Oligotrophy to the UlTra Oligotrophy PACific Experiment (OUTPACE cruise, Feb. 18 to Apr. 3, 2015),” Biogeosciences 14(13), 3207–3220 (2017). [CrossRef]
28. M. Pujo-Pay and P. Raimbault, “Improvement of the wet-oxidation procedure for simultaneous determination of particulate organic nitrogen and phosphorus collected on filters,” Mar. Ecol.: Prog. Ser. 105, 203–207 (1994). [CrossRef]
29. A. Aminot and R. Kérouel, Dosage automatique des nutriments dans les eaux marines : méthodes en flux continu, (Ifremer, 2007).
30. L. A. Currie, “Detection and quantification limits: origins and historical overview,” Anal. Chim. Acta 391(2), 127–134 (1999). [CrossRef]
31. P. Raimbault and G. S. Slawyk, “A semiautomatic, wet-oxidation method for the determination of particulate organic nitrogen collected on filters,” Limnol. Oceanogr.: Methods 36(2), 405–408 (1991). [CrossRef]
32. A. Wong, R. Keeley, and T. Carval, “Argo quality control manual,” Argo Data Management1–50 (2013).
33. C. Schmechtig, H. Claustre, A. Poteau, and F. D’Ortenzio, “Bio-Argo quality control manual for the Chlorophyll-A concentration”, (2018).
34. C. Schmechtig, E. Boss, N. Briggs, H. Claustre, G. Dall’Olmo, and A. Poteau, “BGC Argo quality control manual for particles backscattering”, (2019).
35. C. G. de Boyer Montegut, A. S. Madec, A. Fischer, A. Lazar, and D. Iudicone, “Mixed layer depth over the global ocean: An examination of profile data and a profile-based climatology,” J. Geophys. Res.: Oceans 109(C12), C12003 (2004). [CrossRef]
36. C. Roesler, J. Uitz, H. Claustre, E. Boss, X. Xing, E. Organelli, N. Briggs, A. Bricaud, C. Schmechtig, A. Poteau, F. D’Ortenzio, J. Ras, S. Drapeau, N. Haëntjens, and M. Barbieux, “Recommendations for obtaining unbiased chlorophyll estimates from in situ chlorophyll fluorometers: A global analysis of WET Labs ECO sensors,” Limnol. Oceanogr.: Methods 15(6), 572–585 (2017). [CrossRef]
37. J. Sullivan, M. Twardowski, S. Ronald, J. V. Zaneveld, and C. C. Moore, Measuring optical backscattering in water, Light scattering reviews 7, pp. 189–224. (Springer Kokhanovsky, Berlin, Germany, 2013).
38. N. Bock, F. V. Wambeke, M. Dion, and S. Duhamel, “Microbial community structure in the western tropical South Pacific,” Biogeosciences 15(12), 3909–3925 (2018). [CrossRef]
39. V. Martinez-Vicente, G. Dall’Olmo, G. Tarran, E. Boss, and S. Sathyendranath, “Optical backscattering is correlated with phytoplankton carbon across the Atlantic Ocean,” Geophys. Res. Lett. 40(6), 1154–1158 (2013). [CrossRef]
40. R. M. Letelier, R. Bidigare, D. V. Hebel, M. Ondrusek, C. D. Winn, and D. M. Karl, “Temporal variability of phytoplankton community structure-based on pigment analysis,” Limnol. Oceanogr. 38(7), 1420–1437 (1993). [CrossRef]
41. C. D. Winn, J. Campbell, J. R. Christian, R. M. Letelier, D. V. Hebel, J. E. Dore, L. Fujieki, and D. M. Karl, “Seasonal variability in the phytoplankton community of the north pacific subtropical gyre,” Global Biogeochem. Cycles 9(4), 605–620 (1995). [CrossRef]
42. S. Blain S, S. Bonnet, and C. Guieu, “Dissolved iron distribution in the tropical and subtropical South Eastern Pacific,” Biogeosciences 5(1), 269–280 (2008). [CrossRef]
43. T. Moutin, D. M. Karl, S. Duhamel, P. Rimmelin, P. Raimbault, B. A. S. Van Mooy, and H. Claustre, “Phosphate availability and the ultimate control of new nitrogen,” Biogeosciences 5(1), 95–109 (2008). [CrossRef]
44. R. A. Reynolds, D. Stramski, and B. G. Mitchell, “A chlorophyll-dependent semi analytical model derived from field measurements of absorption and backscattering coefficients within the Southern Ocean,” J. Geophys. Res.: Oceans 106(C4), 7125–7138 (2001). [CrossRef]
45. T. K. Westberry, G. Dall’Olmo, E. Boss, M. J. Behrenfeld, and T. Moutin, “Coherence of particulate beam attenuation and backscattering coefficients in diverse open ocean environments,” Opt. Express 18(15), 15419–15425 (2010). [CrossRef]
46. Y. Huot, A. Morel, M. S. Twardowski, D. Stramski, and R. A. Reynolds, “Particle optical backscattering along a chlorophyll gradient in the upper layer of the eastern South Pacific Ocean,” Biogeosciences 5(2), 495–507 (2008). [CrossRef]
47. M. Barbieux, J. Uitz, A. Bricaud, E. Organelli, A. Poteau, C. Schmechtig, B. Gentilli, G. Obolensky, E. Leymarie, C. Penkerc’h, F. D’Ortenzio, and H. Claustre, “Assessing the variability in the relationship between the particulate backscattering coefficient and the chlorophyll a concentration from a global Biogeochemical-Argo database,” J. Geophys. Res.: Oceans 123(2), 1229–1250 (2018). [CrossRef]
48. A. de Verneil, L. Rousselet, A. M. Doglioli, A. A. Petrenko, and T. Moutin, “The fate of a southwest Pacific bloom: gauging the impact of submesoscale vs. mesoscale circulation on biological gradients in the subtropics,” Biogeosciences 14(14), 3471–3486 (2017). [CrossRef]
49. T. Wagener, C. Guieu, R. Losno, S. Bonnet, and N. Mahowald, “Revisiting atmospheric dust export to the Southern Hemisphere ocean: Biogeochemical implications,” Global Biogeochem. Cycles 22(2), GB2006 (2008). [CrossRef]
50. D. J. McGillicuddy, L. A. Anderson, N. R. Bates, T. Bibby, K. O. Buesseler, C. A. Carlson, and D. A. Hansell, “Eddy/wind interactions stimulate extraordinary mid-ocean plankton blooms,” Science 316(5827), 1021–1026 (2007). [CrossRef]
51. C. S. Law, E. M. S. Woodward, M. J. Ellwood, A. Marriner, S. J. Bury, and K. A. Safi, “Response of surface nutrient inventories and nitrogen fixation to a tropical cyclone in the southwest Pacific,” Limnol. Oceanogr. 56(4), 1372–1385 (2011). [CrossRef]
52. H. Berthelot, T. Moutin, S. L’Helguen, K. Leblanc, S. Hélias, O. Grosso, N. Leblond, B. Charrière, and S. Bonnet, “Dinitrogen fixation and dissolved organic nitrogen fueled primary production and particulate export during the VAHINE mesocosm experiment (New Caledonia lagoon),” Biogeosciences 12(13), 4099–4112 (2015). [CrossRef]
53. M. Caffin, H. Berthelot, V. Cornet-Barthaux, A. Barani, and S. Bonnet, “Transfer of diazotroph-derived nitrogen to the planktonic food web across gradients of N2 fixation activity and diversity in the western tropical South Pacific Ocean,” Biogeosciences 15(12), 3795–3810 (2018). [CrossRef]
54. A. C. Martiny, J. A. Vrugt, F. W. Primeau, and M. W. Lomas, “Regional variation in the particulate organic carbon to nitrogen ratio in the surface ocean,” Global Biogeochem. Cycles 27(3), 723–731 (2013). [CrossRef]
55. P. G. Falkowski, “Evolution of the nitrogen cycle and its influence on the biological sequestration of CO2 in the ocean,” Nature 387(6630), 272–275 (1997). [CrossRef]
56. C. M. Moore, M. M. Mills, K. R. Arrigo, I. Berman-Frank, L. Bopp, P. W. Boyd, E. D. Galbraith, R. J. Geider, C. Guieu, S. L. Jaccard, T. D. Jickells, J. La Roche, T. M. Lenton, N. M. Mahowald, E. Marañón, I. Marinov, J. K. Moore, T. Nakatsuka, A. Oschlies, M. A. Saito, T. F. Thingstad, A. Tsuda, and O. Ulloa, “Processes and patterns of oceanic nutrient limitation,” Nat. Geosci. 6(9), 701–710 (2013). [CrossRef]
57. E. D. Galbraith and A. C. Martiny, “A simple nutrient-dependence mechanism for predicting the stoichiometry of marine ecosystems,” Proc. Natl. Acad. Sci. 112(27), 8199–8204 (2015). [CrossRef]
58. R. J. Geider and J. La Roche, “Redfield revisited: variability of C:N:P in marine microalgae and its biochemical basis,” Eur. J. Phycol. 37(1), 1–17 (2002). [CrossRef]
59. S. Duhamel, T. Moutin, F. Van Wambeke, B. Van Mooy, P. Rimmelin, and H. Claustre, “Claustre Growth and specific P-uptake rates of bacterial and phytoplanktonic communities in the Southeast Pacific (BIOSOPE cruise),” Biogeosciences 4(6), 941–956 (2007). [CrossRef]
60. D. W. Menzel and J. H. Ryther, “The composition of particulate organic matter in the western north Atlantic,” Limnol. Oceanogr. 9(2), 179–186 (1964). [CrossRef]
61. A. Paytan, B. J. Cade-Menun, K. McLaughlin, and K. L. Faul, “Selective phosphorus regeneration of sinking marine particles: evidence from 31P-NMR,” Mar. Chem. 82(1-2), 55–70 (2003). [CrossRef]
62. B. Schneider, R. Schlitzer, G. Fischer, and E. M. Nöthig, “Depth-dependent elemental compositions of particulate organic matter (POM) in the ocean,” Global Biogeochem. Cycles 17(2), 1032 (2003). [CrossRef]
63. V. Martinez-Vicente, G. H. Tilstone, S. Sathyendranath, P. I. Miller, and S. B. Groom, “Contributions of phytoplankton and bacteria to the optical backscattering coefficient over the Mid-Atlantic Ridge,” Mar. Ecol.: Prog. Ser. 445, 37–51 (2012). [CrossRef]
64. J. R. Graff, T. K. Westberry, A. J. Milligan, M. B. Brown, G. Dall’Olmo, V. van Dongen-Vogels, K. M. Reifel, and J. B. Behrenfeld, “Analytical phytoplankton carbon measurements spanning diverse ecosystems,” Deep Sea Res., Part I 102, 16–25 (2015). [CrossRef]
65. Z. P. Lee, K. L. Carder, and R. Arnone, “Deriving inherent optical properties from water color: A multi-band quasi-analytical algorithm for optically deep waters,” Appl. Opt. 41(27), 5755–5772 (2002). [CrossRef]
66. P. J. Werdell, B. A. Franz, S. W. Bailey, G. W. Feldman, E. Boss, V. E. Brando, and A. Mangin, “Generalized ocean color inversion model for retrieving marine inherent optical properties,” Appl. Opt. 52(10), 2019–2037 (2013). [CrossRef]
67. H. Loisel, D. Stramski, D. Dessailly, C. Jamet, L. Li, and R. A. Reynolds, “An inverse model for estimating the optical absorption and backscattering coefficients of seawater from remote-sensing reflectance over a broad range of oceanic and coastal marine environments,” J. Geophys. Res.: Oceans 123(3), 2141–2171 (2018). [CrossRef]
