Phosphorus is a critical component of the genetic and energetic machinery of all life and plays key structural roles in most organisms. Recent debate has bolstered the case that phosphorus is not only essential for life on Earth but is also likely central for recognisable biochemistry more broadly. Recent biogeochemical modeling and reconstructions of the evolution of marine phosphate concentrations from Earth’s rock record have bolstered the case that phosphorus has been the ultimate limiting nutrient throughout Earth’s history. It has also been argued that phosphorus would be expected to limit the extent of life on exoplanets where oxygenic photosynthesis has evolved. Therefore, a mechanistic understanding of how the global phosphorus cycle has changed through time is crucial both for a basic understanding of the history of life on our planet and in the development of predictive frameworks for the production and maintenance of exoplanet biosignatures.
In a paper published on the arXiv database at Cornell University Library on 18 February 2020, Drew Syverson of the Department of Geology and Geophysics at Yale University, Christopher Reinhard of Earth and Atmospheric Sciences at the Georgia Institute of Technology, the Alternative Earths Team at the NASA Astrobiology Institute, and the NASA Nexus for Exoplanet System Science, Terry Isson, also of the Department of Geology and Geophysics at Yale University, and of the Faculty of Science & Engineering at the University of Waikato (Tauranga), Cerys Holstege and Joachim Katchinoff, again of the Department of Geology and Geophysics at Yale University, Benjamin Tutolo of the Department of Geosciences at the University of Calgary, Barbara Etschmann and Joël Brugger of Earth, Atmosphere & Environment at Monash University, and Noah Planavsky, again of the Department of Geology and Geophysics at Yale University, and the Alternative Earths Team at the NASA Astrobiology Institute, consider the posibility of the presence of bioavailable free phosphorus being available on an anoxic early Earth or exoplanet as a result of the weathering of mafic oceanic crust, and develop an experimental proceedure to test this idea.
Continental weathering, an important long-term carbon dioxide sink regulating planetary climate, is also the only significant source of phosphorus to the modern oceans. Submarine weathering of basaltic oceanic crust, while also serving as a long term carbon dioxide sink, currently acts a significant removal process of phosphorus from marine systems. For example, roughly 20% of the phosphorus entering the oceans today is removed through basalt alteration. The remainder is removed in association with the burial of authigenic apatite, organic phosphorus, and iron oxides in marine sediments. Marine phosphate concentrations are controlled through time by the interplay between the efficiency of phosphorus burial and the magnitude of the phosphorus source(s). It has been proposed that anoxic and iron-rich (ferruginous) oceans, which were widespread on the early Earth led to enhanced phosphate scavenging through adsorption onto iron oxide minerals formed near the oxygenated ocean-atmosphere interface and through the precipitation of reduced iron-phosphate minerals directly from seawater.
Phosphorus removal during basalt alteration may have also been enhanced in Earth’s past, given that it is likely that oceanic crust weathering played a more significant role in weathering and carbon dioxide sequestration prior to the emergence of continents above sea level and the proliferation of land. However, previous work has neglected the possibility that dissolved phosphorus may become liberated into seawater during marine weathering of oceanic crust in the absence of dissolved oxygen, as a natural result of limited iron oxidation and subsequent phosphorus scavenging. Although mid-ocean ridge basalts typically do not contain igneous apatite, phosphorus substitutes for silicon in primary silicate minerals and this phosphorus may be released during submarine basalt alteration. If operative, this process would reshape our view of the evolution of the phosphorus cycle on Earth Plants in terrestrial ecosystems and would become an important component of attempts to predict planetary phosphorus cycling on habitable exoplanets. Syverson et al. provide direct experimental support for the idea that basalt weathering under anoxic marine conditions can be a significant source of bioavailable phosphorus to aqueous systems.
The experiments utilised dissolved silicon dioxide enriched in the isotope Silicon²⁹ as a tracer to directly and accurately correlate the extent of primary silicate mineral dissolution with the amount of phosphate mobilized into seawater upon reaction with natural submarine basalt under anoxic conditions. Two sets of experiments are presented, one utilising fresh submarine basalt and one in which the fresh submarine basalt was treated with a reductive dissolution procedure designed to remove pre-existing iron oxides associated with partial oxidation during recovery from the seafloor. Mineral characterisation of the reactant basalt is combined with time-series changes in dissolved phosphate and Silicon²⁹ enriched silicon dioxide to allow determination of the efficacy of basalt weathering as a source of critical nutrients under the anoxic conditions characteristic of the early Earth and reducing exoplanet waterworlds.
Phosphorus has two valencies, phosphorus³⁺ and phosphorus⁵⁺. Phosphorus³⁺ compounds will typicallt form when phosphorus is exposed to oxygenising elements, such as halogens, and can subsequently form free phosphorus³⁺ ions if such compounds are disolved in water. However. phosphorus⁵⁺ already bound into phosphate ions is considered thermodynamically stable under most conditions; it will not break down into phosphorus³⁺ and oxygen when disolved in water without considerable additional energetic input. The reduced dissolved phosphorus³⁺ ion is therefre not considered in Syverson et al.'s study, since the prescribed conditions of the experiments dictate that phosphorus⁵⁺ species, specifically phosphate, are thermodynamically stable and that phosphorus derived from the dissolution of primary silicate minerals of the reactant basalt exists as phosphorus⁵⁺ .
Oxygenated basalt weathering experiments were also conducted in order to demonstrate the removal and retention of dissolved phosphate as a consequence of the formation of iron oxide minerals upon reaction, providing a comparative analysis of phosphate mobility with the anoxic experiments.
Two sets of long term basalt-seawater alteration experiments were conducted for over 2000 hours to better quantify the mobility of dissolved inorganic phosphate under a range of dissolved oxygen concentrations: Firstly anoxic experiments, (BA-1) and (BA-2), where it was expected that phosphate adsorption would be limited due to the lack of iron oxide mineral formation, both of which were conducted at 25°C. Secondly oxygenated experiments, (BA-3), (BA-4), and (BA-5), with concentrations of oxygen at present atmospheric level, in which Syverson et al. demonstrate the process of phosphate adsorption onto iron oxides is associated with oxidative alteration of basalt. Experiments (BA-3), (BA-4), and (BA-5) were conducted at 75, 50, and 15°C, respectively. All of the experiments were conducted at temperatures elevated with respect to bottom seawater but within the range of temperatures associated with carbonate precipitation in subseafloor upon chemical weathering of the oceanic crust.
The anoxic experiments were prepared, initiated, and maintained under anoxic conditions inside an anaerobic chamber with an atmosphere composed of 95% nitrogen and 5% helium. Each anoxic basalt alteration experiment was conducted in a gas-tight 250 mL Pyrex reactor with two gas-tight sampling ports, allowing for direct time-series sampling of the reactor solution without termination of the experiment and for keeping the reaction system closed to external environmental changes. Additionally, each reactor was wrapped in aluminum foil to shield from ultraviolet radiation, effectively preventing photo-oxidation of dissolved iron liberated from the dissolution of reactant basalt. Commencement of each experiment began by combination of the reactant basalt with the synthetic seawater and immersion into a temperature-regulated circulating oil bath set at 25°C. Solution samples were taken with time to assess the change in the composition of reactant seawater solution over the course of approximately 1500 hours. A typical sampling mass of solution for each sampling session can range between 4 - 8 grams. The first sample, through the sampling line of the reactor, is used as a 'bleed' to effectively wash out the sampling tube with experimental solution. The second sample is taken for cations/anions and pH analysis. A third sample may be taken for additional sample solution for additional analysis. In exchange with experimental solution taken from the reactor, an equivalent volume of 95% nitrogen and 5% helium atmosphere enters the experimental reactor. After the sampling session, both the intake and extraction valves are closed to keep the experimental system closed to any external changes in the chamber throughout thereaction progress.
The reactant solution for each anoxic experiment, (BA-1) and (BA-2), was prepared with deoxygenated ultrapure water and designed to have concentrations of dissolved components to simulate seawater, but with elevated concentrations of dissolved Silicon²⁹ enriched silicon dioxide. The pH was buffered at 6.5 to effectively preclude changes in pH as a consequence of basalt dissolution and acidity produced from secondary mineral formation. The reactant basalt was recovered from near the axis of the Juan de Fuca Ridge and was separated from the basalt glass and the crystalline fraction, where the crystalline fraction was powdered to achieve maximum reactive surface area to enhance reaction progress, sieved to achieve a distribution between 5-50 μm in grain size. A fraction of the crystalline powdered basalt reactant was treated with a citrate-bicarbonate-dithionite reductive dissolution step prior to commencement of the experiments. The citrate-bicarbonate-dithionite method specifically removes pre-existing iron-oxides from the natural reactant basalt, which may have a significant effect on the mobility of phosphate liberated from basalt upon alteration under anoxic conditions. The citrate-bicarbonate-dithionite method does not react with and or dissolve the primary reactant basalt silicate minerals. Thus, the primary phospherous, which resides as a trace element within the primary silicate minerals will remain unreactive (as lattice-bound phosphorus⁵⁺ within silicates) upon citrate-bicarbonate-dithionite treatment of the reactant basalt. Further, if any of the phosphorus⁵⁺ associated with the secondary iron oxide minerals were to become reduced to phosphorus³⁺ he highly soluble phosphorus³⁺ would have dissolved within the citrate-bicarbonate-dithionite leachate solution. Once the citrate-bicarbonate-dithionite procedure was terminated, the citrate-bicarbonate-dithionite treated basalt minerals were immediately filtered, cleaned by washing with copious amounts of ultrapure water, and then dried under an anoxic nitrogen atmosphere, with the result that any liberated phosphorus³⁺ would not have been measured by Syverson et al.'s procedure as being liberated during primary basalt dissolution.
The two experiments, basalt with and without the citrate-bicarbonate-dithionite treatment, (BA-1) and (BA-2), respectively, were designed to have an elevated water/rock ratio (approximately 250 g of solution to 3.8 g of basalt), where the mass balance of silica dioxide in the experimental system is dominated by reactant basalt. Specifically, the Silicon²⁹ enriched silicon dioxide spike introduced with the anoxic synthetic seawater was utilized to quantify the degree of silicate dissolution within basalt with reaction progress. The change in the abundance of Silicon²⁹ relative to Silicon²⁸ between the dissolved and solid reservoirs (seawater and basalt) was monitored with time, where it was expected that the Silicon²⁹/Silicon²⁸ ratio of the solution will decrease with increasing degrees of basaltic silicate dissolution, ultimately approaching the natural Silicon²⁹/Silicon²⁸ ratio (about 5%) due to mass balance constraints in the experimental system. This approach was advantageous since the change in the Silicon²⁹/Silicon²⁸ ratio in the solution was not affected by secondary mineral formation effects in contrast to observed changes in the total concentration of dissolved silicon dioxide.
Three oxygenated experiments, (BA-3), (BA-4), and (BA-5 )were conducted by reacting ground crystalline basalt (~20-250 μm grain size) with a synthetic seawater solution that is enriched in dissolved phosphate relative to natural bottom seawater to illustrate the effect of adsorption onto pre-existing and incipiently formed iron oxide surfaces upon alteration of the reactant basalt. For each experiment, eight identical 30 mL Teflon reactors were designed to have identical water/rock ratios (10 g of seawater solution: 1 g of basalt) and all have a headspace (approximately 20 mL) composed of modern atmosphere. The oxygenated reactant seawater was prepared to have a solution chemistry which closely resembles the composition of modern seawater.These particular experiments did not implement the Silicon²⁹ enriched silicon dioxide spike in the experimental solution but rather used natural abundance reagent-grade silicon dioxide. In addition, these particular experiments did not use a pH buffer but rather allowed the pH to change as a consequence of mineral dissolution and secondary mineral formation. Each reactor was maintained at 75, 50, and 15°C, in (BA-3), (BA-4), and (BA-5), respectively, in a shaking water bath, to maintain the experiment at a constant temperature throughout the course of basalt alteration. Over the course of approximately 4000 hours, individual reactors were terminated sequentially with reaction progress to determine the change in composition of the reactant seawater and to monitor changes in mineral speciation and element distribution of the altered basalt relative to the reactant material.
The enriched dissolved Silicon²⁹ enriched silicon dioxide tracer approach was utilised in Syverson et al.'s study since the change in Silicon²⁹/Silicon²⁸ ratio of the reactant seawater is a reflection only of the dissolution of reactant basalt and not by secondary mineral formation processes. The initially elevated dissolved silicon dioxide concentration of the anoxic experiments, (BA-1) and (BA-2), was designed to prevent rapid mineral dissolution artifacts during reaction initiation, while the observed changes in the dissolved silicon dioxide concentration over time is a result of secondary mineral formation. Indeed, the latter is precisely why Syverson et al. monitored the Silicon²⁹/Silicon²⁸ ratio of the solution rather than solely the dissolved silicon dioxide concentrations, because the Silicon²⁹/Silicon²⁸ tracer is not affected by secondary mineral formation while the dissolved net silicon dioxide mass balance is reflective of dissolution and precipitation effects. Through conservative mixing relationships and by implementation of time-series changes in the concentration of dissolved silicon dioxide and the associated Silicon²⁹/Silicon²⁸ ratio compared with the statistically constant composition of reactant basalt, the total amount of basaltic silicate dissolution can be quantified.
Characterisation of the reactant crystalline basalt, recovered from the axis of the Juan de Fuca Ridge, by electron microprobe analysis demonstrated that phosphate is present in primary silicates and secondary minerals at trace levels, between 0.01-0.4 percent by weight, and electron microprobe maps indicate that phosphorus is also concentrated within secondary minerals along the reaction rims around primary silicates. Electron microprobe imaging and X-ray fluorescence maps coupled with iron X-ray absorption spectroscopy measurements demonstrate that the reaction rims of the primary silicates are composed of secondary iron oxide minerals that are intermixed with clay minerals. The citrate-bicarbonate-dithionite reductive dissolution treatment, specifically designed to remove pre-existing iron oxides associated with the partial oxidation of the reactant basalt at the seafloor, indicates that phosphate and other components highly associated with iron oxides, such as manganese, vanadium, and nickel, are concentrated in the leachate. Dissolved aluminium is also concentrated in the leachate, suggesting that the dissolution of iron bearing clays also occurred upon the citrate-bicarbonate-dithionite leach of the reactant basalt. The combination of these results suggests that the observed differences in the mobility of dissolved phosphate with reaction progress between the anoxic experiments, (BA-1) and (BA-2), is likely attributable to the presence of pre-existing iron oxide minerals associated with the reactant basalt. To circumvent this effect on dissolved phosphate mobility, the anoxic experiment (BA-1) used reactant basalt that had pre-existing secondary iron oxide minerals removed by the citrate-bicarbonate-dithionite reductive dissolution treatment prior to commencement of the experiment.
Electron microscopy maps of aluminium, iron and phosphorus from grains derived from the ground basalt sampled from the Juan de Fuca Ridge used in the experimental study. Phosphorus is concentrated along areas of the altered grains that are concentrated in iron, which is attributed to a mixture secondary iron oxide minerals and clays. Syverson et al. (2020).
The oxidation state of the experimental system determines the fate of iron, either dissolved in solution or as aniron oxide mineral, and is critical with respect to the mobility and bioavailability of phosphate liberated from primary silicates upon partial dissolution. In the oxygenated basalt weathering experiments, (BA-3), (BA-4), and (BA-5), the mobility of dissolved iron and phosphate was limited, where dissolved iron, ultimately derived from the dissolution of primary silicate minerals, remained below the detection limit while the initially elevated concentration of dissolved phosphate decreased with reaction progress as a consequence of adsorption onto pre-existing and incipiently formed iron oxide minerals. Interestingly, the quantitative removal of dissolved phosphate is not complete as a consequence of the surface sites of the iron oxide minerals becoming increasingly saturated with respect to other competing adsorbed dissolved species, such as silicon dioxide.
The most significant difference between the anoxic and oxygenated experiments is the increase in concentration of dissolved iron and phosphate into the synthetic anoxic seawater, in stark contrast to the oxygenated experiments, where the mobility of dissolved iron is limited due to quantitative oxidation as iron-oxide minerals and the effective removal of dissolved phosphate by adsorption. Comparison between the separate anoxic experiments, (BA-1) (with citrate-bicarbonate-dithionite) and (BA-2) (without citrate-bicarbonate-dithionite), suggests that pre-existing iron oxide minerals, formed from partial alteration of primary basalt minerals by seawater, have retained a fraction of adsorbed phosphate prior to the experiment. Further, phosphate and the trace redox-sensitive metals, vanadium and manganese, are all highly concentrated in the citrate-bicarbonate-dithionite leachate, consistent with the high affinity of these elements for adsorption and incorporation within iron oxide minerals, respectively. These observation may help elucidate the observed differences in dissolved phosphate and iron mobility between (BA-1) and (BA-2), as shown by the enhanced release of these components into solution upon commencement of the reaction for experiment (BA-2).
The change in the Silicon²⁹/Silicon²⁸ ratio for experiments (BA-1) and (BA-2) are consistent with each other, where the proportion of disolved Silicon²⁹ in both decrease from approximately 100% to values between 11% and 14% due to the dissolution of reactant basalt and mixing of isotopically natural silicon dioxide into the enriched Silicon²⁹ enriched silicon dioxide-bearing synthetic seawater solution. The consistency between the experiments lends confidence, despite the dynamic changes in the bulk concentration of the dissolved components, magnesium, calcium, and silicon dioxide with reaction progress, that the change in the Silicon²⁹/Silicon²⁸ ratio is reflective only of primary mineral dissolution. Interestingly, both experiments asymptotically approach an elevated Silicon²⁹ value of 11-13%, relative to what is predicted by silicon dioxide mass balance constraints of the system, where it is expected for the Silicon²⁹/Silicon²⁸ in solution to ultimately approach the natural abundance of Silicon²⁹ of 5%. This limited mixing of silicon dioxide derived from basalt into the synthetic seawater is likely due to the formation of a passivating layer of secondary minerals on the primary minerals in the basalt, effectively hindering continued reaction.
Experiment (BA-1) demonstrates a 1:1 linear relationship with the calculated amount of total phosphate liberated from basalt into solution with respect to the observed changes in total phosphate with reaction progress. These time-series data, together with thermodynamic constraints, demonstrate that iron-phosphorous minerals, such as vivianite, do not reach a critical saturation state required for nucleation and dissolved phosphate concentrations are not limited by adsorption partitioning onto clays upon primary silicate mineral dissolution. Experiment (BA-2) however, which did not utilize the reductive dissolution pre-treatment, demonstrates significant scatter of phosphae time-series data relative to what is expected for phosphate released from Silicon²⁹/Silicon²⁸ mass balance calculations, indicating an additional source of phosphate to the experimental solution from the basalt reactant, likely attributed to chemical exchange of the reactant solution with the adsorbed species associated with pre-existing alteration product.
Both anoxic basalt weathering experiments demonstrated significant mobility of both dissolved iron and phosphate with continued reaction progress. The dissolved Silicon²⁹/Silicon²⁸ ratio systematically decreased with reaction progress in both experiments, indicating release of phosphate to seawater upon partial dissolution of primary silicate minerals within the basalt substrate. These results are in stark contrast to the oxygenated basalt weathering experiments, in which dissolved iron, derived from primary silicate minerals, remained below the detection limit while phosphorus was either removed from solution (despite high initial dissolved phosphorus levels) or remained at steady state levels throughout reaction progress as a consequence of adsorption onto pre-existing and incipiently formed iron oxide minerals. This suggests an important dichotomy between oxygenated and anoxic deep oceans in the mobilisation of phosphorus to bioavailable forms during submarine basalt weathering.
The experimental data can be used to evaluate the potential of submarine basalt weathering under anoxic conditions to serve as a source of bioavailable phosphate to the deep ocean by using the timeseries Silicon²⁹/Silicon²⁸ data to quantify the total amount of atmospheric carbon dioxide that would be consumed upon dissolution of primary silicate minerals in basalt. Syverson et al. assume a silicon dioxide/Alkalinity ratio indicative of the composition of tholeiitic basalt, represented here for simplicity as enstatite (magnesium-silica oxide), which would react with carbon dioxide to form magnesium carbonate and silicon dioxide. By combining the time-series changes in Silicon²⁹/Silicon²⁸ with the total dissolved silicon dioxide of the experimental solution, Syverson et al. use mass balance to quantify the total carbon dioxide consumed throughout the reaction progress. Specifically, for both anoxic experiments, (BA-1) and (BA-2), the time-series changes in the dissolved concentration of silicon dioxide and Silicon²⁹/Silicon²⁸ ratio are coupled with the silicon dioxide concentration and isotopic composition of basalt through conservative mixing relationships to calculate the amount of released silicon dioxide into solution from basalt upon dissolution.The total carbon dioxide consumed is determined through stoichiometric relations in the reaction and the calculated total silicon dioxide released upon basalt dissolution.
Syverson et al.'s experimental results indicate that the total phosphate/total carbon dioxide value characteristic of submarine basalt weathering under anoxic conditions is similar to that estimated for modern weathering of the continental crust. This suggests that submarine basalt weathering under anoxic conditions should in many cases be roughly similar in its effectiveness at exporting bioavailable phosphorus during carbon dioxide consumption as the weathering of continental crust above sea level.
To gain insight into the large-scale implications of their experimental results, Syverson et al. used their derived total phosphate/total carbon dioxide value as an input parameter for a global phosphate=carbon dioxide-oxygen steady state mass balance model for a water-rich silicate planet on which basalt alteration is the primary carbon dioxide sink; such that the long-term carbon cycles are buffered entirely by weathering of the oceanic lithosphere. They envisioned this scenario as being applicable to portions of Earth’s early history and to Earthlike volatile-rich exoplanet ‘waterworlds’ on which oxygenic photosynthesis has evolved. Conceptually, phosphateis released to the ocean as volcanic/metamorphic carbon dioxide is consumed during submarine basalt weathering, and some fraction of this bioavailable phosphate passes through the biosphere while the remainder is scavenged. This in turn leads to organic carbon burial and a release of oxygen to the ocean-atmosphere system. Syverson et al. explored a wide range of carbon dioxide outgassing rates, relative phosphorus mobilities, burial (sedimentary) phosphorus to organic carbon ratios, and phosphorus scavenging efficiencies in order to compute net oxygen production by the biosphere. Because the values of many of these parameters are uncertain, Syverson et al. used a stochastic approach in order to compute a statistical distribution for possible global biospheric oxygen fluxes.
Schematic of the biogeochemical model. A specified volcanic flux contributes carbon dioxide to the ocean-atmosphere system. Some fraction of this carbon dioxide input is removed via low-temperature ocean crust alteration, which releases inorganic bioavailable phosphorus (Pi). This bioavailable phosphorus can be recycled through the biosphere and buried in marine sediments (green), which releases oxygen to the atmosphere, or can be scavenged by iron-bearing mineral phases in the ocean interior (red). In the model, the fraction of bioavailable phosphorus that is removed by scavenging and the dynamics of internal phosphorus recycling (e.g., the overall ratio between phosphorus burial in sediments and oxygen release) are denoted by the scavenging efficiency of bioavailable phosphate in the ocean interior and rthe global net carbon/phosphorus, respectively. Syverson et al. (2020).
The parameters of the model are: (1) the volcanic carbon dioxide outgassing rate, which at steady state must be balanced by carbon dioxide removal through either submarine weathering or photosynthetic uptake; (2) the mobility ratio derived from Syverson et al.'s experiments (total phosphate/total carbon dioxide), which describes the molar ratio between carbon dioxide consumed and phosphate released during submarine weathering; (3) the scavenging efficiency of bioavailable phosphate in the ocean interior; and (4) the global net carbon/phosphorus ratio for material buried from the oceans (in effect corresponding to a ratio between phosphorus buried and oxygen released to the ocean-atmosphere system).
Results of Syverson et al.'s simulation indicate that fluxes of phosphate from oceanic crust weathering can support a wide range of biospheric oxygen release rates. However, biospheric oxygen fluxes are generally high across a wide range of parameter space unless volcanic/metamorphic carbon dioxide outgassing is assumed to be very low. Syverson et al.'s results strongly suggest that biospheres sustained entirely by bioavailable phosphorus released during submarine basalt weathering under anoxic conditions, including that of the earliest Earth, are potentially capable of generating extremely high biogenic gas fluxes that rival or exceed even those of the modern Earth.
Studies of Earth system evolution and conceptual models used to forecast the emergence and maintenance of biosignatures on volatile-rich exoplanets have neglected the differences in hydrothermal phosphorus cycling between oxic and anoxic systems, and have instead implicitly invoked conditions under which phosphorus is effectively scavenged through adsorption onto iron oxide minerals formed from the hydration and oxidation of primary silicates in submarine basalt. Syverson et al.'s results stand in strong contrast to this prevailing conceptual model, and thus provide impetus to revisit mechanistic models for Earth’s early oxygen cycle and the factors regulating the oxygen cycles of volatile-rich silicate planets more generally. In particular, it will be important for future work to establish the ocean-atmosphere oxygen ‘threshold’ above which oxygenation of the deep oceans attenuates bioavailable phosphorus fluxes by initiating widespread iron oxidation within the ocean interior.
Intriguingly, existing model results indicate the possibility that atmospheric oxygen could be present at abundances that would potentially be remotely detectable without fully ventilating the deep oceans, with the implication that Earth-like terrestrial planets with water inventories greater than that of Earth could be very promising targets in the search for exoplanet biosignatures. Syverson et al.'s experimental results provide strong evidence that fluxes of bioavailable phosphorus on terrestrial planets dominated by submarine basalt weathering under anoxic conditions are likely to be very robust and in some cases may surpass those of even the modern Earth system. Anoxic submarine weathering should thus be considered an important component of the large-scale redox balance of terrestrial planets.
See also...
Follow Sciency Thoughts on
Facebook.