Dissolved organic phosphorus concentrations in the surface ocean controlled by both phosphate and iron stress

Dissolved organic phosphorus (DOP) has a dual role in the surface ocean as both a product of primary production and as an organic nutrient that fuels primary production and nitrogen fixation, especially in oligotrophic gyres. Although poorly constrained, the geographic distribution and environmental controls of surface ocean DOP concentrations influence the distributions and rates of primary production and nitrogen fixation in the global ocean. Here we pair DOP concentration measurements with a metric of phosphate stress, satellite-based chlorophyll a concentrations and a satellite-based iron stress proxy to explore their relationship with upper 50 m DOP stocks. Our results suggest that phosphate and iron stress work together to control surface ocean DOP concentrations at basin scales. Specifically, upper 50 m DOP stocks decrease with increasing phosphate stress, while alleviated iron stress leads to either surface DOP accumulation or loss depending on phosphate availability. Our work extends the relationship between DOP concentrations and phosphate availability to the global ocean, suggests a linkage between marine phosphorus cycling and iron availability and establishes a predictive framework for DOP distributions and their use as an organic nutrient source that supports global ocean fertility. Production and consumption of dissolved organic phosphorus in the surface ocean is controlled by the interplay between phosphate and iron stress, according to global analyses of the distribution of marine nutrients.

), is scarce [4][5][6][7][8][9][10] . Estimates from the marine ecosystem component of the Community Earth System Model indicate that global marine net primary production (NPP) and dinitrogen (N 2 ) fixation rates are ~8% and ~33% higher, respectively, and better match the observed rates when DOP is included as an assimilative phosphorus source 11 , which was also necessary to match the large-scale gradients in surface ocean bulk DOP concentrations. Like dissolved organic carbon (DOC) and dissolved organic nitrogen (DON), DOP is a product of primary production 12 . However, a clear understanding of both the distribution of DOP in surface waters as well as the controls on those distributions is limited by the relatively small number of DOP concentration measurements in the global ocean 11 . A mechanistic framework to understand the controls on surface ocean DOP concentrations would thus improve model-based estimates of the rates and distributions of marine NPP, N 2 fixation and ultimately the biological pump.
Given that DOP is both produced and consumed by photosynthetic organisms, we explored the relationships between DOP distributions and metrics of primary production (that is, DOP production), PO 4 3− stress (that is, DOP consumption) and iron stress, which can limit primary production. In addition, iron has recently been identified as a co-factor in a version of the enzyme that is responsible for DOP utilization by phytoplankton, that is, alkaline phosphatase 10,13,14 , and thus iron availability may also affect DOP consumption. To test these relationships, we paired surface ocean DOP concentration measurements ( Fig. 1) with monthly satellite-based measurements of chlorophyll a concentrations, climatological 'excess PO 4 3− ' or 'P*' values 15 evaluated in surface waters (<5 m) and a climatological, satellite-based iron stress proxy (NPQ-corrected φ sat values, see below) 16 (Figs. 1 and 2). Although small but notable seasonal changes in DOP concentrations have been found at the ALOHA and BATS stations 3,17 we assume that these DOP concentrations ( Fig. 1) represent steady-state conditions and perform a basin-scale analysis of their distributions. We find that the upper 50 m DOP stocks, which include multiple mixed-layer DOP concentration measurements, are significantly positively correlated with surface P* values (R 2 = 0.28, P < 0.00000001) (Fig. 1b). Here, higher P* values correspond to lower PO 4 3− stress, consistent with stoichiometric biomass demands for nitrogen and phosphorus 18 , as well as PO 4 3− concentration thresholds for DOP utilization by phytoplankton 2 . In addition to PO 4 3− stress, iron stress also plays a significant but more complicated role in regulating surface DOP concentrations and can lead to either surface DOP accumulation under enhanced iron stress or consumption under alleviated iron stress (Fig. 2). On the basis of these relationships, we present a conceptual model of the factors that control surface ocean DOP concentrations.

Global regions of net DOP accumulation and loss
The primary source of dissolved organic matter to the open ocean is marine primary production, and DOC and DON accumulate in regions with elevated productivity 12,[19][20][21][22] . Similarly, we find that upper 50 m DOP stocks are significantly positively correlated with satellite-based measurements of chlorophyll a concentration on the GO-SHIP P18-2016 and BIOSOPE cruises 23 in the eastern Pacific Ocean as well as the Gulf of Mexico (GOM2019 cruise) (Figs. 1 and 2 and Table 1). Correspondingly, the eastern North Pacific, eastern South Pacific and the Gulf of Mexico have the highest mean surface ocean DOP concentrations in our dataset, averaging 0.34 ± 0.07 µM, 0.23 ± 0.07 µM and 0.23 ± 0.07 µM, respectively (Extended Data Fig.  1), and represent regions of net DOP production. By contrast, the North Atlantic, western North Pacific and western South Pacific surface ocean DOP and chlorophyll a concentrations are significantly negatively correlated ( Fig. 2 and Table 1). In these samples, when Dissolved organic phosphorus concentrations in the surface ocean controlled by both phosphate and iron stress Zhou Liang 1 ✉ , Robert T. Letscher 2 and Angela N. Knapp 1 Dissolved organic phosphorus (DOP) has a dual role in the surface ocean as both a product of primary production and as an organic nutrient that fuels primary production and nitrogen fixation, especially in oligotrophic gyres. Although poorly constrained, the geographic distribution and environmental controls of surface ocean DOP concentrations influence the distributions and rates of primary production and nitrogen fixation in the global ocean. Here we pair DOP concentration measurements with a metric of phosphate stress, satellite-based chlorophyll a concentrations and a satellite-based iron stress proxy to explore their relationship with upper 50 m DOP stocks. Our results suggest that phosphate and iron stress work together to control surface ocean DOP concentrations at basin scales. Specifically, upper 50 m DOP stocks decrease with increasing phosphate stress, while alleviated iron stress leads to either surface DOP accumulation or loss depending on phosphate availability. Our work extends the relationship between DOP concentrations and phosphate availability to the global ocean, suggests a linkage between marine phosphorus cycling and iron availability and establishes a predictive framework for DOP distributions and their use as an organic nutrient source that supports global ocean fertility.
chlorophyll a is high, DOP concentrations are low, indicating that these regions are not associated with net DOP production but that instead DOP is used as an assimilative phosphorus source sustaining productivity. Thus, the North Atlantic, western North Pacific and western South Pacific appear to be net sinks for DOP and have the lowest observed basin-mean concentrations of 0.10 ± 0.07 µM, 0.12 ± 0.02 µM and 0.14 ± 0.04 µM, respectively (Extended Data Fig.  1). Below we explore the interdependence of surface ocean DOP concentrations, primary productivity, PO 4 3− stress (P*) and iron stress.
surface ocean DOP primarily controlled by PO 4

3−
The primary control on surface ocean DOP concentrations is PO 4 3− stress, as gauged by P*, with upper 50 m DOP stocks in all data sets significantly positively correlated with P* (Fig. 1b). This positive relationship is robust whether calculating P* using upper   4 3− stress is considered to be the primary control on surface ocean DOP distributions.

iron stress modifies DOP accumulation and loss
Whereas higher PO 4 3− stress leads to enhanced DOP consumption on the global scale, alleviated iron stress can promote either DOP production or consumption regionally. Remotely sensed fluorescence quantum yield (φ sat ) data have been used as a proxy for iron stress experienced by phytoplankton after correcting for non-photochemical quenching (NPQ) 16,[27][28][29][30] . Here we use climatological, NPQ-corrected φ sat data as a relative measure of the iron stress experienced by phytoplankton 16 , with higher NPQ-corrected φ sat values corresponding to higher iron stress. In the GO-SHIP P18-2016 and BIOSOPE 23 eastern Pacific Ocean samples, NPQ-corrected φ sat and upper 50 m DOP stocks are significantly negatively correlated (Table 1 and Fig. 2), which we interpret as reflecting enhanced DOP production when iron stress is alleviated in high-chlorophyll a upwelling regions (Table 1 and Fig. 2). A significant negative correlation between upper 50 m DOP stocks and NPQ-corrected φ sat is also found on the West Florida Shelf in the Gulf of Mexico (GOM2019 cruise) ( Table 1 and Fig. 2). On the global scale, the eastern Pacific Ocean appears unique as a region of net DOP production in upwelling-associated, relatively low PO 4 3− and iron stress surface waters, with DOP subsequently advected west towards oligotrophic gyres experiencing higher PO 4 3− and iron stress, ultimately contributing to DOP loss within the western Pacific Ocean.
In contrast to the eastern Pacific Ocean and Gulf of Mexico, basins with net DOP consumption exhibit significant positive correlations between upper 50 m DOP stocks and NPQ-corrected φ sat (Table 1 and Fig. 2). Specifically, in samples from the GO-SHIP P06-2017 cruise in the subtropical South Pacific, the AMT17, AMT14 and 36N cruises 2 from the Atlantic Ocean, and the KH12-3 cruise 31 from the western North Pacific, upper 50 m DOP stocks increase with increasing iron stress (Fig. 2). We interpret the higher DOP stocks in these waters with higher iron stress as reflecting the iron limitation of primary productivity, and thus decreasing demand for DOP as an assimilative phosphorus source, and/or iron limitation of alkaline phosphatase activity 13,14 , thus limiting the ability of phytoplankton to use DOP. The low DOP concentrations (<0.15 µM) in the North Atlantic and western Pacific Ocean are found in regions with modest chlorophyll a concentrations (that is, typically >0.1 mg chlorophyll a per m 3 ), low P* values (<0.1 µM) and low NPQ-corrected φ sat values (Fig. 2), which is consistent with DOP use by phytoplankton as an assimilative phosphorus source when iron is available and PO 4 3− is scarce. The North Atlantic in particular receives high rates of dust deposition 32 , and the western South Pacific may receive significant hydrothermal iron fluxes 33 , lowering the iron stress in these regions. These regions are also associated with high rates of N 2 fixation 34-37 , which may be due to certain diazotrophs (for example, Trichodesmium spp.) being particularly well adapted to utilizing DOP when PO 4 3− is scarce 4,6,26,38 . In addition, DOP appears to accumulate in regions where surface currents converge and iron stress is relatively high, thus limiting productivity. Specifically, in the convergence zone of the eastern South Pacific (Fig. 2j and Extended Data Fig. 4), we observed elevated DOP concentrations (~0.3 µM) in waters with high NPQ-corrected φ sat , low P* and low chlorophyll a concentrations (Fig. 2j-l). The same scenario was found in the surface convergence zone of the South Atlantic, where the DOP concentration was higher than at the gyre boundary ( Fig. 2m and Extended Data Fig. 4). We suggest that DOP will accumulate in the surface convergence zone of the eastern North Pacific (Extended Data Fig. 4) as well, which is   another region with high iron stress, low P* and low chlorophyll a. Further sampling would test this hypothesis. NPQ-corrected φ sat data as a measure of iron stress have been evaluated in NO 3 − -replete surface waters 39,40 , with one recent study finding evidence of elevated NPQ-corrected φ sat in low-NO 3 − surface waters that corresponded with molecular markers of iron stress 30 . Since we evaluate φ sat in environments with a range of NO 3 − concentrations, we underscore the novel nature of our correlations of NPQ-corrected φ sat with DOP stocks in low nutrient surface waters, which will best be further investigated via future field studies characterizing the phytoplankton community composition, their photo-physiology, and molecular markers of iron and light stress 30,41,42 . However, we also note that evaluation of NPQ-corrected φ sat data using a range of NPQ correction schemes yields the same sense and statistical significance of correlation with DOP stocks (Supplementary Figs. 1 and 2), although the absolute value of NPQ-corrected φ sat should be used with caution. Photoacclimation adds uncertainty to φ sat data 28 ; however, using 16 year climatological NPQ-corrected φ sat data temporally averages some of this variability, and most observations investigated here come from the low to middle latitudes, where photoacclimation has a smaller effect 28 . We also acknowledge the reduced signal-to-noise ratio of φ sat in waters with a low normalized fluorescence line height (nFLH) (typically <0.003 mW cm −2 µm −1 sr −1 ) [43][44][45] , which are marked with triangle symbols in Fig. 2. We find no significant changes in the relationships between DOP stock and NPQ-corrected φ sat when these samples are excluded. Nevertheless, we emphasize the evolving nature of remote-sensing-based ocean colour metrics 16,[27][28][29][30] that are best validated through direct observations using multiple tools, in this case the composition of the phytoplankton community, photo-physiology and molecular markers of iron and light stress.

conceptual model of controls on surface ocean DOP
According to the observed relationships between upper 50 m DOP stocks, surface chlorophyll a concentrations and PO 4 3− and iron stress, we propose a simple conceptual model of the factors that influence DOP distributions in the surface ocean (Fig. 3). The four quadrants in Fig. 3 correspond to different PO 4 3− and iron stress regimes that reflect the corresponding role of DOP as either a product of or substrate for primary productivity in specific ocean regions. On the global scale, DOP loss is enhanced under elevated PO 4 3− stress, which increases to the right along the x axis in Fig.  3, with surface DOP concentrations decreasing to the right and increasing to the left. In addition, iron stress, increasing along the y axis, promotes both DOP production and consumption, depending on PO 4 3− stress. As discussed above, the eastern Pacific Ocean is a highly productive region with low PO 4 3− stress, resulting in net DOP production and accumulation ( Figs. 1 and 2), and is represented by the left pink-coloured quadrants of Fig. 3. Iron stress further influences the magnitude of DOP accumulation under low PO 4 3− stress. When both iron and PO 4 3− stress are alleviated, such as in surface waters overlying ODZs, enhanced primary production will lead to significant net DOP accumulation, so that DOP has a 'production' signature. By contrast, the upper-left light-pink quadrant reflects regions in the eastern Pacific Ocean that exhibit more muted net DOP accumulation due to surface ocean convergence of DOP produced in 'upstream' regions and to the lower, iron-limited rates of primary productivity locally. Global surface ocean regions with the lowest DOP concentrations are associated with high PO 4 3− stress and low iron stress, such as the North Atlantic, western North Pacific and western South Pacific (Fig. 1). These regions are represented by the dark-blue, lower-right quadrant in Fig. 3 and correspond to regions of net DOP loss that we interpret as reflecting the use of DOP as an assimilative phosphorus source that sustains productivity, perhaps particularly by diazotrophs. Finally, the upper-right, light-blue quadrant corresponds to regions experiencing relatively balanced degrees of enhanced PO 4 3− and iron stress, potentially including the South Atlantic, although observation gaps exist. The South Atlantic receives relatively low atmospheric dust inputs 32 , and the lack of significant rates of water column denitrification and/or anammox in the eastern portion of the basin leaves low P* values in the surface  waters ( Figs. 1 and 2). Here, net changes in DOP distributions are small, suggesting the potential for co-limitation of primary productivity by PO 4 3− and iron in the region. The linear regression model fitting the upper 50 m DOP stocks versus P* values (Fig. 1b) predicts a surface DOP concentration at the BATS site (46 ± 22 nM) that is similar to observations (~60 nM (ref. 3 )). However, the predicted surface DOP concentration at Station ALOHA (146 ± 31 nM) is lower than observations (~200 nM (ref. 46 )), with the difference potentially reflecting iron limitation of DOP consumption, which is not accounted for in this correlation (Fig. 1). Given the higher atmospheric dust fluxes to the North Atlantic relative to the North Pacific near Hawaii, it is reasonable to expect that iron limitation plays a smaller role in controlling DOP distributions at the BATS site. Although the simple linear relationship does not capture all the processes that influence surface ocean DOP concentration (R 2 = 0.28), the predictions reflect the observed, basin-scale differences in surface DOP concentrations (Extended Data Fig. 5). Meanwhile, the modelled global surface ocean DOP distribution is improved when including P* and NPQ-corrected φ sat data with or without the chlorophyll a concentration as predictors using three machine-learning algorithms (Extended Data Table 1).

implications
Our observations demonstrate significant, basin-scale differences in correlations of upper 50 m DOP stocks with climatological, inorganic-nutrient concentration ratios and remote-sensing products of surface ocean chlorophyll a concentration and iron stress. On the basis of these observations we present a predictive conceptual model for the controls on surface ocean DOP concentrations. Net DOP production is observed in regions with elevated P* values (lower PO 4 3− stress) and relatively low iron stress, consistent with elevated rates of productivity and low pressure on the DOP pool as an assimilative phosphorus source. Net DOP consumption is apparent in regions with P* values of <0.1 µM and relatively low iron stress, suggesting that phytoplankton growth is limited by PO 4 3− availability and not iron in these regions. This is consistent with emerging work evaluating the role of nitrogen, phosphorus and iron limitation in different ocean basins 47 . Although a link between PO 4 3− stress and DOP use as an organic nutrient by phytoplankton has been established in the Atlantic Ocean 2,14 , our analysis identifies a potential role for iron availability in modulating DOP accumulation versus consumption in global ocean surface waters. The enhanced role for DOP use as an organic nutrient in oligotrophic waters requires that DOP accumulates elsewhere, with iron availability potentially serving in a role similar to that in the regulation of intrabasin N 2 fixation rates 48 . Such a predictive framework for the distribution of surface ocean DON concentrations is lacking, suggesting a decoupling between these two organic nutrients. This mechanistic model of surface DOP concentration distributions in the ocean provides a testable framework for both observational and modelling work, and can help to identify conditions where DOP acts as an organic nutrient source augmenting rates of NPP and N 2 fixation. Indeed, the region with the greatest net DOP loss, that is, 0.2 µM between 130° W and 80° E in the South Pacific, is consistent with high rates of N 2 fixation 36,37,48,49 , indicating that DOP is probably an important phosphorus source that fuels N 2 fixation in this region where PO 4 3− concentrations are low (<0.1 µM) and iron stress is reduced. In addition, surface ocean DOP consumption enhances ocean-to-atmosphere fluxes of methane 50 , a potent greenhouse gas, and a predictive understanding of where DOP consumption occurs may improve methane flux estimates. Finally, our observations linking marine DOP cycling with iron availability extend the previously described coupling of the marine iron, carbon and nitrogen cycles to include phosphorus as well.

Online content
Any methods, additional references, Nature Research reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/ s41561-022-00988-1.