Modeling p N 2 through Geological Time: Implications for Planetary Climates and Atmospheric Biosignatures

Abstract

Nitrogen is a major nutrient for all life on Earth and could plausibly play a similar role in extraterrestrial biospheres. The major reservoir of nitrogen at Earth's surface is atmospheric N₂, but recent studies have proposed that the size of this reservoir may have fluctuated significantly over the course of Earth's history with particularly low levels in the Neoarchean—presumably as a result of biological activity. We used a biogeochemical box model to test which conditions are necessary to cause large swings in atmospheric N₂ pressure. Parameters for our model are constrained by observations of modern Earth and reconstructions of biomass burial and oxidative weathering in deep time. A 1-D climate model was used to model potential effects on atmospheric climate. In a second set of tests, we perturbed our box model to investigate which parameters have the greatest impact on the evolution of atmospheric pN₂ and consider possible implications for nitrogen cycling on other planets. Our results suggest that (a) a high rate of biomass burial would have been needed in the Archean to draw down atmospheric pN₂ to less than half modern levels, (b) the resulting effect on temperature could probably have been compensated by increasing solar luminosity and a mild increase in pCO₂, and (c) atmospheric oxygenation could have initiated a stepwise pN₂ rebound through oxidative weathering. In general, life appears to be necessary for significant atmospheric pN₂ swings on Earth-like planets. Our results further support the idea that an exoplanetary atmosphere rich in both N₂ and O₂ is a signature of an oxygen-producing biosphere. Key Words: Biosignatures—Early Earth—Planetary atmospheres. Astrobiology 16, 949–963.

1. Introduction

Life as we know it is implausible without nitrogen. It is an essential major nutrient for all living things because it is a key component of the nitrogenous bases in the nucleic acids that store, transcribe, and translate genetic information, a necessary ingredient of the amino acids constituting the proteins responsible for most cellular catalysis and at the core of the ATP molecule that is the principal energy transfer agent for biological metabolism. As nitrogen's cosmic abundance is only slightly less than that of carbon and oxygen and because it condenses at moderate distances in circumstellar disks, it should have been available to other potential exoplanetary biospheres. However, N₂ gas, the most likely nitrogen species near planetary surfaces in the habitable zone, is nearly inert at standard conditions because of the very strong triple bond in N≡N. Lightning can break this bond and combine nitrogen with other atmospheric species, but on Earth this process has been relatively inefficient (Borucki and Chameides, 1984; Navarro-González et al., 2001). Much more significant for removing N₂ from Earth's atmosphere is microbial N₂ fixation to ammonium, a reaction catalyzed by the enzyme nitrogenase. Today, most of this fixed nitrogen is returned to the atmosphere via the biogeochemical nitrogen cycle, a series of microbially modulated redox reactions that ultimately transform organic nitrogen back to gaseous N₂. Thus, life's demand for nitrogen regulates its atmospheric abundance.

Though N₂ is not a greenhouse gas itself, its atmospheric partial pressure affects planetary environmental conditions. Higher N₂ pressure can enhance greenhouse warming by pressure broadening the absorption bands of such gases as water vapor, CO₂, CH₄, and N₂O (Goldblatt et al., 2009). Hence, the partial pressure of N₂ indirectly influences surface temperature and thus habitability. Moreover, a low total atmospheric pressure of all gases combined weakens the cold-trap for water vapor at the tropopause (Wordsworth and Pierrehumbert, 2014). Where N₂ is a major atmospheric constituent, a drop in pN₂ can make the tropopause cold-trap leaky (Zahnle and Buick, 2016), allowing water vapor into the upper layers of the atmosphere where it can either remain as vapor, perhaps increasing overall greenhouse warming (Rind, 1998; Solomon et al., 2010; Dessler et al., 2013) (but for a contrary view see Huang et al., 2016), or freeze as high-altitude ice clouds in polar regions, warming the high latitudes and making planetary climate more equable (Sloan and Pollard, 1998). Thus, planetary habitability is dependent, at least in part, on atmospheric nitrogen levels.

Though we know next to nothing about the evolution of the biogeochemical nitrogen cycles on other planets, we now have a better-resolved picture of the behavior of nitrogen through Earth's history (Ader et al., 2016; Stüeken et al., 2016). It seems that (1) microbial nitrogen fixation evolved very early in Earth's history such that a nitrogen crisis for the primordial biosphere was averted (Stüeken et al., 2015; Weiss et al., 2016), (2) the partial pressure of atmospheric N₂ has fluctuated through time to a greater degree than previously anticipated (Som et al., 2016), (3) an aerobic nitrogen cycle arose before the Great Oxidation Event (GOE) at ∼2.35 Ga (Garvin et al., 2009; Godfrey and Falkowski, 2009), (4) during the mid-Proterozoic aerobic and anaerobic nitrogen cycling was spatially separated under low-oxygen conditions (Stüeken, 2013; Koehler et al., 2016), and (5) a modern nitrogen cycle with widespread aerobic activity did not arise until the late Neoproterozoic (Ader et al., 2014). The main constraints on these developments were evidently biological evolution and redox changes in Earth's surface environments, principally the oxygenation state of the atmosphere and ocean (Stüeken et al., 2016). Other Earth-like planets may have evolved along somewhat similar pathways with respect to nitrogen cycling, provided that they also originated Earth-like life. If so, then atmospheric swings in pN₂ may be a common feature of terrestrial inhabited planets.

In the present study, we investigated the diversity of terrestrial planetary nitrogen cycles by modeling the evolution of Earth's atmospheric N₂ reservoir. We then perturbed the model to examine several ahistorical extreme scenarios that could arise on Earth-like exoplanets, defined here as planets with a silicate rock mantle and iron core (empirically <1.6 Earth radii in size, Rogers, 2015), an orbit within the conservative limits of the habitable zone (Kopparapu et al., 2013), and a similar volatile content to Earth. This conservative definition prescribes a high-molecular-weight atmosphere dominated by N₂, CO₂, and H₂O rather than H₂ (cf. Pierrehumbert and Gaidos, 2011; Seager, 2013). After exploring a range of variables, we concluded that some combinations of N₂ abundance with other gases could act as extraterrestrial biosignatures, others could be “false positives,” and yet others may indicate that a planet is uninhabited. Overall, our results suggest that an anaerobic biosphere can greatly facilitate the removal of large amounts of N₂ from a planetary atmosphere.

2. Model Setup

2.1. Biogeochemical nitrogen box model

We used the isee Stella software to construct a box model of the global biogeochemical nitrogen cycle (Fig. 1), tracking total nitrogen. Boxes included the atmosphere, pelagic marine sediments deposited on oceanic crust, continental marine sediments deposited on continental shelves and in epeiric seas, continental crust, and the mantle. Continental marine sediments were defined to become part of the continental crust after 100 million years and from there onward were subjected to metamorphism, which returns nitrogen to the atmosphere. Similarly, we let pelagic marine sediments accumulate for 100 million years before they were subjected to subduction, metamorphism, and volcanism. These timescales were based on the observation that C/N ratios in sedimentary rocks older than about 100 million years become more variable with higher average values (Algeo et al., 2014). The two sinks of nitrogen from the atmosphere were burial in pelagic sediments and burial in continental sediments with proportions of ∼1:25 (Berner, 1982). The sources of atmospheric nitrogen were volcanic and metamorphic degassing of subducted pelagic sediments, metamorphism of continental crust, oxidative weathering of continents, and mantle outgassing. Denitrification was not explicitly included as a source, because we did not track the marine nitrogen reservoir.

FIG. 1.

Schematic of the nitrogen model for Earth. All fluxes except for burial of nitrogen from the atmosphere into sediments are dependent on the reservoir size. Fluxes with dashed arrows are implemented as time-variable, except in our “base model” (Fig. 2a).

The model was run in 1-million-year time-steps from 4.5 Ga to the present. Differential equations listed in the Supplementary Material were solved with the Euler method (Supplementary Material is available online at www.liebertonline.com/ast). The chosen step size is much lower than the residence time of nitrogen in our modeled reservoirs (>100 million years), which eradicates computational artifacts that can result in mass imbalances. The initial abundances of nitrogen in the mantle, continental crust, and sediments were set such that the concentrations were the same, assuming that any disequilibrium in concentrations observed today is due to biogeochemical overprinting. The rock masses and modern nitrogen inventories were taken from the work of Johnson and Goldblatt (2015). The initial abundance of atmospheric N₂ is unknown. We tuned the model such that the final atmospheric N₂ abundance after 4.5 billion years equaled the modern amount of 2.87·10²⁰ mol, defined as one times present atmospheric nitrogen (PAN) (Johnson and Goldblatt, 2015).

A more detailed description of how rate constants were derived is given in the Supplementary Material. Following the work of Berner (2006), nitrogen burial was parameterized through biomass burial (i.e., organic carbon). Although this approach neglects the origin and radiation of biological nitrogen metabolisms over Earth's history (Stüeken et al., 2016), it is preferred because (a) it avoids major uncertainties about metabolic rates in deep time and (b) it is sufficient for tracking the total nitrogen sink from the atmosphere. As further discussed below (Section 4.1), additional nitrogen burial through adsorption on clay minerals is negligible compared to the organic nitrogen flux into sediments. Using this approach, we calculated modern nitrogen burial from the modern volcanic CO₂ outgassing flux of 6·10¹⁸ mol/Myr (Marty and Tolstikhin, 1998), assuming that 22% (f _org = 0.22) of volcanogenic CO₂ is buried as organic carbon in the absence of land plants (pre-Devonian) (Krissansen-Totton et al., 2015) and that the C/N ratio of post-diagenetic marine biomass is approximately 10 (Godfrey and Glass, 2011; Algeo et al., 2014). This gave a nitrogen burial flux of 1.32·10¹⁷ mol/Myr, equal to ∼1.5% of modern biological N₂ fixation (8.64·10¹⁸ mol/Myr) (Galloway et al., 2004). We then modulated this flux in deep time in three different ways. First, we took into account secular trends in f _org as inferred from the carbon isotope record (“F_org model”) (Krissansen-Totton et al., 2015). Uncertainties in this record, resulting from potentially underrepresented carbonate reservoirs (Bjerrum and Canfield, 2004; Schrag et al., 2013) and the variance in δ¹³C values at any given time point, are discussed below. Second, we assumed a gradual decline in CO₂ outgassing from the Hadean to the modern, following the parameterization of Canfield (2004, Eq. 2) (“F_org + Heatflow model”). By mass balance, higher CO₂ outgassing in the earlier Precambrian implies higher burial fluxes of biomass and with it nitrogen. Uncertainties and caveats of this approach are discussed below. Third, we tested for the effects of additional 2-fold increases in CO₂ input, and hence, nitrogen burial, during superplume events (Abbott and Isley, 2002) (“F_org + Heatflow + Superplume model”). This model did not include potential N₂ addition through enhanced volcanism, which we will discuss separately. In all these models, oxidative weathering was implemented as a function of pO₂ through time (Lyons et al., 2014) with a reaction order of 0.5, that is, proportional to (pO₂)^0.5 (Chang and Berner, 1999; Bolton et al., 2006).

In a separate set of models, we tested hypothetical scenarios for Earth-like planets without any biosphere and an anoxic atmosphere, or with a completely anaerobic biosphere that does not experience oxygenation events. Our definition of Earth-like planets includes an orbit within the habitable zone where liquid water can be stable at the surface, as well as a rocky composition and a volatile inventory similar to those of Earth (Section 1). This definition is admittedly limited, but it allows us to identify with greater confidence a subset of parameters that can play a critical role in the history of the nitrogen cycle. We assumed abiotic nitrogen burial was driven by NH₄ ⁺ adsorption on clay minerals after abiotic N₂ reduction in the atmosphere. Rates were taken from the work of Stüeken (2016) for two extreme end-members corresponding to high and low estimates of abiotic N₂ fixation rates for Archean Earth and seawater pH of 5 and 8, respectively (see Supplementary Material S1.1). For illustration, these burial fluxes are equal to 1.4·10⁻¹ and 1.7·10⁻⁵ times the modern nitrogen burial described above. We did not test abiotic burial under oxic conditions because in that case NH₄ ⁺ production would likely be low and nitrogen burial trivial. For the “biotic anoxic” scenario, we used nitrogen burial rates equal to 1–10% of the modern biological N₂ fixation rate. In both the abiotic and biotic anoxic model, oxidative weathering was switched off.

Rate constants for oxidative weathering, metamorphism, erosion, subduction, volcanism, and outgassing were calculated as the ratio of modern fluxes to modern reservoir sizes (see Supplementary Material for details). We further performed simple order-of-magnitude estimates of catastrophic events, such as impacts and large volcanic eruptions, to test whether these could affect our overall conclusions (Section 3.3).

2.2. Modeling effects on greenhouse gases and global climate

We selected the most extreme end-members of our Earth model to determine potential effects on Earth's climate. As shown below, the timescales over which variations in pN₂ can occur exceed those of the carbonate-silicate feedback cycle (Walker et al., 1981). Hence, changing pN₂ alone is unlikely to cause rapid climate changes, because pCO₂ can rapidly re-adjust to keep surface temperatures more or less steady. We therefore decided to calculate the required changes in pCO₂ that would counterbalance the effects of varying pN₂. We used an iterative approach to determine the pCO₂ necessary to maintain a 278 K globally averaged surface temperature (T _GAT), equivalent to the estimated globally averaged temperature during the last glacial maximum. This limit was chosen because geological evidence suggests that the Archean was cool (Hren et al., 2009; Blake et al., 2010) but not permanently glaciated (Young, 1991; de Wit and Furnes, 2016). As input values, we used the changes in nitrogen abundance shown in Section 3 (Fig. 2) and changes in solar luminosity from 3.5 to 2.4 Ga. To do this, we used a 1-D radiative convective model that was recently used to calculate habitable zone boundaries (Kasting et al., 1984, 1993; Kopparapu et al., 2013) and characterize the surface temperature of a hypothetical Archean Earth with a global hydrocarbon haze (Arney et al., 2016). We chose a surface albedo of A _surf = 0.32, which is a tuning parameter used to reproduce the temperature of modern Earth (Kopparapu et al., 2013; Arney et al., 2016). Our atmospheric composition consisted of only N₂, CO₂, H₂O, and CH₄, where the pN₂ was taken from our model output (Table 1) and CH₄ was fixed at pCH₄ = 0.001 (fCH₄, fCO₂, fN₂, and the total surface pressure P ₀ were adjusted self-consistently for changes in pCO₂). The H₂O profiles were calculated using a relative humidity profile assuming a surface relative humidity of 80% (Manabe and Wetherald, 1967). The solar luminosity at 3.5 and 2.4 Ga (0.769 and 0.831) was found through interpolation of Table 2 in the work of Bahcall et al. (2001). We focused on the pCO₂ required to maintain a T _GAT > 278 K, noting that significantly lower temperatures (T _GAT < 273 K) may allow some open oceans, but a 3-D model would be needed to capture their additional complexities (see Supplementary Material S5 for limitations of our approach). Table 1 shows a summary of our results. We also indicate the surface temperature if pCO₂ was not adjusted to maintain T _GAT = 278 K, but was maintained at the pCO₂ value at 3.5 Ga, and if both the pCO₂ and solar luminosity were maintained at the 3.5 Ga values. Table 2 shows similar results as Table 1 but for a more permissive initial and final T _GAT of 273 K. This illustrates the sensitivity of the required pCO₂ adjustment to compensate for pN₂ drawdown as a function of the choice of reference temperature.

FIG. 2.

Reconstructing nitrogen burial over Earth's history. Arrows mark pN₂ value inferred for proxies (Marty et al., 2013; Som et al., 2016). (a) Base model. N burial is held constant, calculated as the product of volcanic CO₂ outgassing, the pre-Devonian organic burial fraction (f _org) of 0.22, and the inverse of the Redfield C/N ratio of 10. (b) F_org model. N burial is modulated by secular changes in f _org as inferred from the carbon isotope record (Krissansen-Totton et al., 2015). (c) F_org + Heatflow model. CO₂ outgassing is assumed to have declined gradually from the Hadean to the modern with decreasing heatflow from Earth's interior, as described by Canfield (2004). Carbon burial and with it nitrogen burial is changed in proportion. Mantle outgassing is modulated in the same way. (d) F_org + Heatflow + Superplumes model. Additional pulses of CO₂ outgassing are assumed during intervals of superplumes as recorded in the rock record (Abbott and Isley, 2002). Solid black line = using best estimates for all parameters; dashed lines = most extreme uncertainty interval if all parameters are off in the same direction (excluding uncertainties in Redfield ratio and modern CO₂ outgassing, Supplementary Material S3); gray shaded area = more plausible uncertainty interval with narrower range of values for the metamorphic rate constant (most sensitive variable, see Supplementary Material S3).

Table 1.

Climate Response to Changes in Atmospheric N₂

Line	Age (Ga)	Rel. solar luminosity	N₂ (PAN)	pN₂ (bar)	pCO₂ (bar)	pCH₄ (bar)	P₀ (bar)	T_GAT (K)
F_org + Heatflow model:
1	3.5	0.769	0.61	0.476	0.046	0.001	0.523	278.0
2	2.4	0.769	0.42	0.328	0.046	0.001	0.375	274.7
3	2.4	0.831	0.42	0.328	0.046	0.001	0.375	282.4
4	2.4	0.831	0.42	0.328	0.018	0.001	0.346	278.0
F_org + Heatflow + Superplumes model, lower limit:
1	3.5	0.769	0.54	0.421	0.056	0.001	0.480	278.0
2	2.4	0.769	0.08	0.062	0.056	0.001	0.119	264.8
3	2.4	0.831	0.08	0.062	0.056	0.001	0.119	270.6
4	2.4	0.831	0.08	0.062	0.160	0.001	0.223	278.0

Input data are the age, the corresponding relative solar luminosity taken from Bahcall et al. (2001), atmospheric N₂ (PAN) calculated from our box model, and an assumed constant background level of 1000 ppmv CH₄. Partial N₂ pressure (pN₂) was calculated as the product of total N₂ in units of PAN and the modern pN₂ of 0.78 bar. Output parameters are T _GAT (global average surface temperature), pCO₂, and P ₀ (total average surface pressure, i.e., sum of all gases). Line 1 = starting conditions at 3.5 Ga; line 2 = changing PAN with constant luminosity and constant pCO₂ (taken from line 1); line 3 = changing PAN and luminosity with constant pCO₂; line 4 = changing PAN, luminosity, and pCO₂. Parameters are calculated such that the T _GAT at (1) and (4) converges to 278 K.

Table 2.

Climate Response to Changes in Atmospheric N₂

Line	Age (Ga)	Rel. solar luminosity	N₂ (PAN)	pN₂ (bar)	pCO₂ (bar)	pCH₄ (bar)	P₀ (bar)	T_GAT (K)
F_org + Heatflow model:
1	3.5	0.769	0.61	0.476	0.020	0.001	0.497	273.2
2	2.4	0.769	0.42	0.328	0.020	0.001	0.348	270.6
3	2.4	0.831	0.42	0.328	0.020	0.001	0.348	278.4
4	2.4	0.831	0.42	0.328	0.005	0.001	0.333	273.2
F_org + Heatflow + Superplumes model, lower limit:
1	3.5	0.769	0.54	0.421	0.024	0.001	0.446	273.2
2	2.4	0.769	0.08	0.062	0.024	0.001	0.087	261.1
3	2.4	0.831	0.08	0.062	0.024	0.001	0.087	266.2
4	2.4	0.831	0.08	0.062	0.118	0.001	0.181	273.2

Same as Table 1, but for a fiducial T _GAT of 273.2 K at lines (1) and (4). Differences from Table 1 are in bold.

3. Results

3.1. Nitrogen burial constrained by carbon burial

We tested four different models for the evolution of nitrogen burial through geological time (Fig. 2). In our base model (Fig. 2a), nitrogen burial was held constant at its modern flux. Each subsequent iteration incorporates an additional parameter to the N burial record and, thus, shows additional atmospheric N drawdown. Sensitivity tests are presented in Section S3. The majority of the uncertainty range illustrated in Fig. 2 derives from uncertainties about rates of nitrogen loss during continental metamorphism.

3.1.1. F_org model

In the first test, the base model was modified by scaling nitrogen burial as a function of f _org (organic carbon burial fraction). This did not significantly change the evolutionary trend of the atmospheric nitrogen reservoir predicted by the base model, as f _org has been relatively invariant through geological time (Fig. 2b). During the largest burial event indicated by the carbon isotope record, the Lomagundi Event at ca. 2.3–2.1 Ga, f _org temporarily exceeds 0.3 (Precambrian baseline ∼0.15–0.2). However, even this increase in organic burial alone appears to be insufficient to significantly deplete the atmospheric N₂ reservoir by more than ∼0.02 PAN. This model shows a slow depletion of the atmospheric nitrogen reservoir during the Archean (from 1.0 to 0.81 PAN) and relatively constant atmospheric nitrogen during the Proterozoic (range 0.78 to 0.82 PAN). Modern atmospheric nitrogen levels are not attained until late in the Phanerozoic.

3.1.2. F_org + Heatflow model

When we scale nitrogen burial proportionally to the amount of CO₂ outgassing (Canfield, 2004), our model shows a significant drawdown of atmospheric nitrogen during the Archean, reaching a minimum of 0.44 PAN ( = 0.35 bar N₂) in the earliest Paleoproterozoic, immediately prior to the GOE (Fig. 2c). Fixed nitrogen is principally buried and stored in continental sediments and continental crust, which reach a maximum of 1.7× the modern continental reservoir size at this time (Section S2). During the GOE, the atmospheric nitrogen reservoir rebounds due to enhanced oxidative weathering of the continents, but this rebound stops after the GOE, and atmospheric nitrogen remains low at 0.59–0.65 PAN ( = 0.47–0.51 bar N₂) until the Neoproterozoic Oxidation Event (NOE). Atmospheric N₂ rapidly rises during the Neoproterozoic and Paleozoic in response to a further enhancement of oxidative weathering with the second rise of O₂. The later Phanerozoic shows a slow gradual increase in atmospheric N₂. As discussed below, we consider this model the most plausible mechanism for explaining a Neoarchean pressure minimum of 0.23 ± 0.23 bar as determined from basaltic amygdales (Som et al., 2016).

3.1.3. F_org + Heatflow + Superplumes model

Scaling up N burial by a factor of 2 during plume intervals (Abbott and Isley, 2002) resulted in a small but noticeable effect on the atmospheric nitrogen reservoir (Fig. 2d). The effect is most pronounced during the longest superplume event in the geological record, which is thought to have lasted from 2.775 to 2.696 Ga. Atmospheric nitrogen is drawn down to 0.42 PAN ( = 0.33 bar N₂) by 2.69 Ga in the wake of this event, which is also consistent with basalt amygdale paleobarometry estimates for 2.7 Ga (Som et al., 2016).

To test the plausibility of greater burial enhancement factors, we explored the range from 1× to 10× burial enhancement during plume intervals. A 10× burial enhancement factor for superplume intervals caused the atmospheric nitrogen reservoir to become entirely depleted in the Neoarchean (0.0 PAN at 2.71 Ga), which is implausible because it would have driven the biosphere to near extinction by removing a major nutrient, or could have induced a global glaciation, for which there is no geological evidence at this time. We thus deemed 10× too extreme of a burial enhancement factor. Values less than 10× and greater than 1× are plausible, but given available data, it is difficult to derive accurate constraints. It is also quite possible that different plume events affected organic burial by different factors. Furthermore, as noted in Section 3.3, nitrogen outgassing from the mantle may have been more pronounced during superplume events, offsetting the effect of enhanced carbon burial. This effect was not considered in our model. All these complications make assessing the impact of plumes on the atmospheric nitrogen reservoir rather difficult. Still, it is worth noting that, even with a 2× burial enhancement, effects on the atmospheric nitrogen reservoir from plume events are small due to their short duration.

3.2. Atmospheric nitrogen evolution on abiotic and anoxic worlds

3.2.1. Abiotic anoxic planets

In the hypothetical scenario of a completely abiotic Earth-like planet in the habitable zone of another star with plausible abiotic nitrogen burial rates (Stüeken, 2016), the atmospheric nitrogen reservoir increased slightly from 1.0 PAN to 1.14–1.22 PAN after 4.5 billion years, with no periods of atmospheric N₂ drawdown (Fig. 3a). Catastrophic events such as plumes would have little effect on such a planet because, as further discussed below (Section 3.3.2), most of the nitrogen liberated by such events on Earth is likely sourced from the crust that has been enriched in nitrogen due to biological burial. On an abiotic planet, the crust would be relatively depleted in nitrogen. The sensitivity of these trends to changes in surface and deep Earth nitrogen fluxes is given in the Supplementary Material (Section S4). Overall, we find no plausible mechanism that could cause large swings in pN₂, apart from the possibility of atmospheric erosion (e.g., Mars, Section 4.3) or freeze-out of N₂ on planets far outside the habitable zone (e.g., Pluto). Although this modeled scenario is hypothetical, it emphasizes the potential importance of life for the evolution of the global nitrogen cycle.

FIG. 3.

Hypothetical evolution of atmospheric N₂ on extrasolar planets. (a) Abiotic anoxic planet where nitrogen burial is driven by NH₄ ⁺ adsorption on clay minerals. High burial rate = 1.5·10¹⁶ mol/Myr, low burial rate = 2.2·10¹² mol/Myr (Stüeken, 2016). (b) Biological anoxic planet without oxidative remineralization and weathering. Nitrogen burial is calculated as the product of the productivity factor and the modern biological N₂ fixation rate on Earth of 8.6·10¹⁸ mol/Myr (Galloway et al., 2004). (c) Biological anoxic planet with extinction of the biosphere upon atmospheric loss. (d) Anoxic biological burial and extinction as in (c) with and without burial on continental crust. The burial flux is set equal to 4% of modern biological N₂ fixation [dashed line in (d) describes the same scenario as dashed line in (c)]. Note change in scale on the x axis.

3.2.2. Biotic anoxic planets

Our results for a hypothetical Earth-like planet with an anaerobic biosphere suggest that, under the right conditions, biological nitrogen drawdown can have a major effect on the evolution of atmospheric pN₂ through time. Nitrogen burial rates equal to 0.1 and 0.04 times the modern N₂ fixation rate sequester atmospheric nitrogen to 0 PAN by 4.1 and 2.9 Ga, respectively (Fig. 3b). The minimum flux required to reach 0 PAN within 4.5 billion years is roughly 0.03 times modern biological N₂ fixation, or 2 times the modern nitrogen burial flux. As in the case of our “F_org + Heatflow” model above, most of this nitrogen is stored in continental sediments and crust. A flux of 0.01 times modern biological N₂ fixation does not completely draw down atmospheric N₂ but leads to a steady state of ∼0.84 PAN (Fig. 3b). A more detailed sensitivity analysis of these simulations is presented in Section S4. When other parameters are set to their most conservative values (i.e., minimizing N₂ sequestration), a fixation flux of ∼0.13 times modern would be needed to draw down atmospheric N₂ to 0 PAN. Again, this scenario is hypothetical, but in comparison to our Earth model, it emphasizes the important influence of atmospheric oxygen on the nitrogen cycle.

It is conceivable that a biosphere would go extinct when the atmospheric N₂ reservoir becomes depleted and triggers a global glaciation, at least if biological N₂ drawdown does not slow down dramatically as pN₂ approaches 0 PAN (e.g., Klingler et al., 1989). Of course, in reality, biogeochemical feedbacks that were not considered in our model may maintain a low, but nonzero, N₂ reservoir in the atmosphere, and the biosphere may not necessarily go extinct as evidenced by extreme glacial events on Earth. We nevertheless explored this case in our model, because it provides an estimate of how fast pN₂ can recover on an anoxic planet in the absence of an Earth-like atmospheric oxygenation event (cf. Fig. 2c, 2d). Our results show that in this case the buried nitrogen would return much more slowly through continental metamorphism and erosion than it does with oxidative weathering. We determined the recovery time of atmospheric pN₂ after it is completely sequestered by switching off biological nitrogen fixation once atmospheric N₂ reached 0 PAN. With nitrogen burial fluxes of 10% and 4% modern N₂ fixation, it takes 2.76 and 2.62 billion years, respectively, for atmospheric nitrogen values to recover to 1.0 PAN, assuming that the biosphere does not recover during that time (Fig. 3c). This is much slower than the increases in pN₂ that occurred in our models over a few hundred million years after the GOE and NOE on Earth, which highlights the linkage between pO₂ and pN₂ that is further discussed below (Section 4.3).

To assess the effects of continental crust (the major nitrogen repository in our models), we ran a separate model where all burial was directed to pelagic sediments; that is, continents were bypassed to mimic a planet without an equivalent of continental crust. Burial fluxes were arbitrarily set to 4% times the modern fixation rate. Under these conditions, pN₂ drawdown to 0 PAN is much slower (Fig. 3d), because nitrogen in pelagic sediments has a much shorter residence time than in continental sediments and crust and is returned relatively rapidly through subduction-zone metamorphism and volcanism. However, once atmospheric N₂ has been sequestered in the mantle, it does not recover, even over a billion-year timescale. Although this model describes a purely hypothetical scenario of plate tectonics in the absence of continents, it demonstrates that a relatively shallow nitrogen repository akin to continental crust with an intermediate residence time between that of pelagic sediments and the mantle greatly facilitates large atmospheric pN₂ swings over hundreds of millions of years, as suggested for Precambrian Earth (Som et al., 2016).

3.3. Catastrophic events

Planets are periodically subject to catastrophic events throughout their history. We performed order-of-magnitude calculations to test whether such events could affect atmospheric N₂ reservoirs and, hence, our overall conclusions. Possible events considered here include asteroid or comet impacts, superplume volcanism as an N₂ source, and large-scale planetary resurfacing.

3.3.1. Impacts

There are three ways impacts could affect the amount of nitrogen in a planetary atmosphere: direct nitrogen addition from the bolide, atmospheric erosion with loss of N₂ to space, and release of buried nitrogen through crustal heating. A variety of asteroids have substantial nitrogen contents. The most volatile rich are carbonaceous chondrites, which have 1235 ppm nitrogen on average (Wasson and Kallemeyn, 1988; Johnson and Goldblatt, 2015). If all nitrogen were released as N₂ upon impact, then a hypothetical 3.2·10²¹ kg of carbonaceous chondrites could contribute 1 PAN. Such a mass is about 0.05% of Earth's mass and is approximately equivalent to the once-proposed total mass of late heavy bombardment material (Wetherill, 1975). Impacts later in the Archean, post-dating the Late Heavy Bombardment, would have been several orders of magnitude smaller than 0.05% of Earth's mass (Johnson and Melosh, 2012); hence this nitrogen source was most likely trivial for atmospheric evolution.

Impact erosion has been proposed as an explanation for the thin atmosphere of Mars (Melosh and Vickery, 1989), but more recent calculations suggest that this mechanism may have been less effective than previously thought (Manning et al., 2009). The effect would further decrease with more massive planets than Mars; it is not generally considered to have been significant on early Earth. It is thus probably not a major nitrogen sink on most habitable Earth-like planets.

Large-scale crustal heating resulting from impacts could add some nitrogen to the atmosphere (Wordsworth, 2016). Current estimates of nitrogen in continental crust suggest a mass of 1.7·10¹⁸ kg, or 0.5 PAN (Goldblatt et al., 2009; Johnson and Goldblatt, 2015). During the late Archean, when atmospheric pN₂ may have been as low as 0.44 PAN (Section 3.1.2), the continental reservoir may have been as large as 1.7 times the present continental reservoir, though nitrogen concentrations in continental crust rocks through time are poorly constrained (Section S2). The nitrogen contained in oceanic crust and lithosphere is relatively minor (∼0.57 to 0.67·10¹⁸ kg N) (Johnson and Goldblatt, 2015). To raise atmospheric pN₂ by more than 0.1 PAN through crustal melting, more than 5% of the continental crust would have needed to melt in the Neoarchean and more than 20% in the modern. We consider this unlikely, because there is no geological evidence of such large-scale crustal melting events. We note, however, that some rock types, such as lower continental crust and the oceanic lithospheric upper mantle, are poorly characterized. There are suggestions that parts of the mantle may contain substantial nitrogen (Li et al., 2015), and if a significant fraction of the mantle experienced melting, then large quantities of nitrogen could be added to the atmosphere.

3.3.2. Superplumes

Superplume events represent another type of mantle melting that may contribute to changes in atmospheric N₂. While typical mantle melts (i.e., MORB) have a very low nitrogen content of ∼1 ppm (Marty, 1995; Marty and Zimmermann, 1999; Johnson and Goldblatt, 2015), sparse data from continental basalts show higher concentrations. Basalts from the Abitibi region have 6 ppm nitrogen (Honma, 1996), Columbia River basalt has 34 ppm (BCR-1, Govindaraju, 1994), and recent Antarctic basalts around 60 to 100 ppm (Greenfield, 1991). Assuming that most nitrogen will degas during eruption, these concentrations suggest that substantial nitrogen could have been released during flood basalt eruptive events.

Fluid in equilibrium with basaltic melt under oxidizing conditions (f _O2 = ΔNNO + 3) has approximately 10⁴ times more nitrogen than the melt itself (Li et al., 2015). Assuming all nitrogen contained in the fluid degasses, an eruption of 3·10¹⁸ kg basalt (equivalent to the Siberian Traps) with ∼30 ppm nitrogen remaining in the basalt suggests a release of 1·10¹⁸ kg nitrogen, or 0.25 PAN. While measurements of nitrogen in flood basalts are quite rare, making this speculative, this simple calculation hints that these events could influence atmospheric N₂ content. However, it is important to note that superplumes in oceanic plates would likely have been much less effective, given that marine basalts tend to have more than 1 order of magnitude less nitrogen than continental flood basalts (see above); the latter likely assimilate and liberate nitrogen from continental crust. An exception may be oceanic superplumes that pass through parts of the mantle that are enriched in nitrogen due to a subduction overprint. Lamproites and lamprophyres, which are volcanic rocks resulting from the melting of enriched mantle, are notably nitrogen-rich (tens of parts per million) (Jia et al., 2003), suggesting that such plume events could potentially liberate large amounts of nitrogen into the atmosphere. Nitrogen could also have been liberated by plumes that sampled the lower mantle, which according to some estimates may be nitrogen-rich (Johnson and Goldblatt, 2015). However, both the nitrogen concentration of the lower mantle and the transport pathways of material from such great depth (>660 km) are highly uncertain, making this mechanism difficult to evaluate. Another complication in predicting the effect of superplumes in deep time comes from the possibility that the redox state of magmas may have increased over time, and under more reducing conditions magmatic nitrogen may have been less volatile (Libourel et al., 2003; Kadik et al., 2011; Roskosz et al., 2013; Mikhail and Sverjensky, 2014).

3.3.3. Planetary resurfacing

If superplume events are extended to a planetary scale, as is suggested to have happened on Venus, even more nitrogen could be released. The area of the Siberian traps, to continue the above example, is 2.5·10⁶ km², or about 1/200^th of the surface of Earth. Multiplying the above estimate for nitrogen released during Siberian Trap volcanism by 200 yields a nitrogen output of 2·10²⁰ kg, or ∼50 PAN. Again, we note that this is a highly speculative estimate, but it does suggest the possibility of substantial additions of nitrogen to the atmosphere via large-scale volcanism on other planetary bodies that contain substantial amounts of nitrogen in the crust. Overall, catastrophic events could have more marked effects on planets where the crust is nitrogen-enriched, which, as noted above (Section 3.1.2, 3.2.2), is more likely to be the case on planets with a large biosphere that transfers atmospheric nitrogen to crustal repositories.

4. Discussion

4.1. The evolution of atmospheric pN₂ on Earth

Although many parameters in the global biogeochemical nitrogen cycle are uncertain and potential reconfigurations of Earth's interior are not taken into consideration for lack of quantitative constraints (Mikhail and Sverjensky, 2014), our results allow us to draw several broad conclusions under the assumption of persistent tectonic cycling through Earth's history, as follows:

(1) The results from our base model (Fig. 2a), where nitrogen burial is held constant through time while oxidative weathering follows atmospheric pO₂, show that the oxygenation of the atmosphere alone could probably not have caused the large swings in atmospheric pN₂ that were suggested by Som et al. (2016). Changes in the atmospheric nitrogen reservoir by more than ∼0.1 PAN most probably require a change in nitrogen burial over time.

(2) Variations in nitrogen burial by up to a factor of 2.9, as inferred from the record of relative organic carbon burial (f _org), are insufficient to cause significant swings of more than ∼0.2 PAN in atmospheric N₂ (Fig. 2b). Even if we use f _org values from the upper end of the uncertainty range (Krissansen-Totton et al., 2015), the atmospheric N₂ reservoir does not drop by more than 0.3 PAN. If f _org was smaller than assumed, due to a greater proportion of carbonate formation in oceanic crust or within sediments (Bjerrum and Canfield, 2004; Schrag et al., 2013), this variable would have even less effect on atmospheric N₂. To reach atmospheric pressures of less than 0.5 bar at 2.7 Ga (Som et al., 2016), while maintaining pressures of 0.5–1.1 bar N₂ at 3.5 Ga (Marty et al., 2013), nitrogen burial must have been markedly higher in the earlier Archean than it is today. There are three possible scenarios to increase nitrogen burial: (i) subduction was more efficient than it is today, and metamorphic devolatilization was suppressed; (ii) nitrogen was buried preferentially relative to carbon; (iii) the absolute organic carbon burial flux was much higher, and with it the burial of carbonate, such that f _org did not change. It is conceivable that subduction was faster in the Archean (option i), but our sensitivity tests (Section S3) show that shortening the residence time of nitrogen in pelagic sediments by a factor of 2 has minimal effects on the atmospheric reservoir (<0.01 PAN), because the bulk of sedimentary nitrogen is recycled via metamorphism. Metamorphic devolatilization may have been enhanced in the Archean when the geothermal gradient was perhaps somewhat higher than today (Condie and Korenaga, 2010; Cartigny and Marty, 2013), but variations in this parameter also have minimal influence on the output of our model (Section S3). The effects of more rapid subduction and enhanced devolatilization may have more or less canceled each other without a net increase of nitrogen drawdown.

Preferential nitrogen burial (option ii) could potentially occur through adsorption of NH₄ ⁺ onto clay minerals. Boatman and Murray (1982) experimentally derived a relationship between the amount of NH₄ ⁺ adsorbed on clay and the dissolved NH₄ ⁺ concentration in solution. For a doubling of the total nitrogen burial flux, the adsorbed concentration would have to match the concentration of organic nitrogen. Shale samples of sub-greenschist metamorphic grade typically have nitrogen concentrations in the range of 100–1000 ppm with C/N ratios around 40, suggesting that most nitrogen is derived from organics (Stüeken et al., 2016). To achieve this concentration through NH₄ ⁺ adsorption alone would require a dissolved NH₄ ⁺ concentration of 3–30 mM, which is 30–300 times higher than the NH₄ ⁺ concentration of the modern Black Sea (∼100 μM, Brewer and Murray, 1973) and 100–1000 higher than modern marine NO₃ ^- levels (∼30 μM, Sverdrup et al., 1942). Such high ammonium abundances are also inconsistent with the nitrogen isotope record, which suggests that N-limited ecosystems dominated by biological N₂ fixation were initiated in the Mesoarchean at 3.2 Ga (Stüeken et al., 2015) and persisted until the GOE at ∼2.35 Ga (Stüeken et al., 2016). A large reservoir of dissolved NH₄ ⁺ should have resulted in isotopic fractionations of up to 27‰ associated with partial NH₄ ⁺ assimilation into biomass (Hoch et al., 1992; Pennock et al., 1996), which is not observed. Moreover, due to the extreme energetic cost of splitting the N≡N triple bond, nitrogen fixation should have been suppressed rather than expressed where ammonium was readily available as a nutrient. Enhanced nitrogen burial through adsorption is further inconsistent with the record of C/N ratios, because significant addition of adsorbed NH₄ ⁺ would require consistently lower C/N ratios in the earlier Precambrian, which is opposite to observations; C/N ratios tend to increase with increasing age due to preferential nitrogen loss during low-grade metamorphism (Stüeken et al., 2016). NH₄ ⁺ adsorption to clays likely did occur during diagenesis, where NH₄ ⁺ in pore waters can become enriched to several millimolar by degradation of organic matter (e.g., Rosenfeld, 1979; Boudreau and Canfield, 1988), but in that case the adsorbed nitrogen is simply returned to the sediment from which it was derived and does not lead to excess nitrogen burial. Enhanced NH₄ ⁺ adsorption in the Archean ocean is therefore unlikely to have caused a drawdown in atmospheric N₂. Instead, we find it more likely that absolute organic carbon burial, and with it organic nitrogen burial, was significantly higher (option iii).

(3) Enhanced volcanic CO₂ outgassing in the earlier Precambrian could explain greater nitrogen burial if accompanied by increased burial of both organic matter and carbonate. The observed constancy of f _org could have been maintained through carbonatization of oceanic crust as a large carbonate sink (Nakamura and Kato, 2004). Organic matter burial was likely concentrated under anoxic waters along continental margins where sedimentation rates were high. Such an enhanced biomass burial flux would have led to an increase in nitrogen burial and can thus explain the observation of low Neoarchean pN₂ (Fig. 2c). Following the formulation of Canfield (2004) for a higher heatflow and proportionally higher volcanic fluxes in deep time, this scenario increases the nitrogen burial flux by 40–80% (depending on the exact age) relative to our early Paleozoic base value, or 3.4–1.9 times above the f _org factor alone, throughout most of the Archean. We note that the assumption of a higher Archean CO₂ flux (Canfield, 2004; Zahnle et al., 2006) has been challenged by studies arguing for a gradual increase in CO₂ outgassing from the Archean into the Proterozoic, concurrent with late continental growth (Holland, 2009; Lee et al., 2016). However, if volcanic CO₂ emissions were lower in the Archean than they are today, absolute carbon burial would have been lower and with it the burial of nitrogen. Low Archean CO₂ fluxes would only be compatible with high nitrogen burial if f _org had been much higher than generally assumed. A high Archean CO₂ flux thus remains the most plausible mechanism in our model to explain a decline in pN₂ from 0.5–1.1 bar at 3.5 Ga to <0.5 bar at 2.7 Ga, as suggested by paleobarometers (Marty et al., 2013; Som et al., 2016). We will therefore proceed with this assumption, noting that the Archean CO₂ flux requires additional constraints to derive a more accurate trajectory for pN₂.

We further note that a higher absolute burial flux of organic carbon would constitute a source of oxygen equivalents that would have needed to be balanced by reductants to maintain anoxic surface conditions in the Archean (Kasting, 2013). Proposed fluxes of possible reductants (H₂, CO, H₂S, Fe²⁺) range over an order of magnitude (reviewed by Zahnle et al., 2006; Holland, 2009) and could therefore plausibly cover the effect of carbon burial. Reductant fluxes may indeed have been higher than previously suggested during the earlier Archean if new evidence for a secular increase in the redox state of Earth's mantle is taken into account (Nicklas et al., 2015). A 2- to 4-fold higher carbon burial flux does therefore not necessarily violate redox balance models.

If we calibrate our model with the results of Som et al. (2016), who inferred an atmospheric pressure of <0.5 bar from the relative sizes of vesicles in basalt flows, then the lower part of our uncertainty range in Fig. 2c is more likely to be correct than the upper part. In this case, our model predicts the lowest atmospheric pressure in the Neoarchean and two stepwise increases across the Paleoproterozoic and Neoproterozoic oxidation events, when oxidative weathering progressively shortens the residence time of nitrogen in continental sediments and crust. Our prediction of ∼0.6 PAN ( = 0.47 bar N₂) in the Mesoproterozoic is testable with further analyses of vesicle sizes in Proterozoic lava flows.

(4) The effect of superplumes is difficult to assess; they could have led to either more rapid nitrogen recycling through crustal melting or slightly enhanced nitrogen drawdown through carbon burial. The balance may further depend on the redox state of magmas, which may have changed over time in favor of progressively more N₂ degassing (Mikhail and Sverjensky, 2014). Overall, reconfigurations of the deep Earth are currently poorly constrained, but these could potentially have significant effects, beyond the scope of our model.

4.2. Climatic effects of atmospheric pN₂ changes in the Archean

Significant changes of >0.1 PAN in our modeled atmospheric N₂ abundances occur over several hundred million-year time scales (Fig. 2c, 2d). Although atmospheric pressure affects the greenhouse efficiency of other atmospheric gases like CO₂ through pressure broadening (Goldblatt et al., 2009) and can therefore theoretically cause changes in surface temperature, these time scales are so long that any resulting temperature change could be balanced by the carbonate-silicate feedback cycle, which has a response time on the order of a few hundred thousand years (Sundquist, 1991). As pN₂ declines, greenhouse warming decreases, causing the planet to cool. However, with lower temperatures, silicate weathering by carbonic acid slows down, which lowers the sink flux of atmospheric CO₂ from the atmosphere (Walker et al., 1981). CO₂ would thus build up and balance the temperature change caused by the drop in pN₂.

Table 1 shows the required response in pCO₂ to our calculated drop in pN₂ in the Archean. These calculations also take into account the increasing solar luminosity, which warms the planet and therefore leads to a steady decrease in pCO₂. In sum, the effect of rising solar luminosity overpowers the effect of declining pN₂ from 3.5 to 2.7 Ga in our nominal model scenario (Fig. 2c); hence pCO₂ would have needed to decrease to maintain a stable surface temperature of 278 K. If pCO₂ did not respond, then surface temperature would increase by about 4 degrees over this time interval due to the increase in solar luminosity, despite the drop in pN₂. It is only in cases of extreme nitrogen burial, that is, at the lower limit of our uncertainty interval in the model with an additional superplume (Fig. 2d, excluding potential effects of crustal melting), that the decline in pN₂ would cause surface temperature to drop by around 7.5 degrees. This drop could have been counterbalanced by a 3- to 5-fold increase in pCO₂. We note that all our calculated values for pCO₂ fall within, or very close to, the range of late Archean CO₂ pressures inferred from the rock record (0.003–0.15 bar at 2.5 Ga and 0.004–0.75 bar around 2.7 Ga, Sheldon, 2006; Driese et al., 2011; Kanzaki and Murakami, 2015). Although these estimates vary widely, this agreement suggests that our model results are not unrealistic. Overall, plausible changes in atmospheric pN₂ as inferred from our model are unlikely to have resulted in massive climatic perturbations. [We note that a requirement for globally averaged temperatures approaching modern values (∼288 K) or higher throughout the Archean would be hard to reconcile with the most restrictive constraints on pCO₂ from paleosols even without the climatic impact of falling pN₂, which is the well-known Faint Young Sun Paradox. For all but our most extreme scenarios, falling pN₂ would only negligibly exacerbate this problem.]

4.3. N₂ in extraterrestrial atmospheres

Geological and potential biological processes on other planets may differ markedly from those on Earth, as might the initial volatile inventory. Our results can therefore only provide qualitative trends, but they may nevertheless serve as useful guidelines in evaluating measurements of atmospheric pN₂ in exoplanetary atmospheres (Schwieterman et al., 2015b). At the very least, our approach allows us to isolate selected variables that have the potential to play a major role in the evolution of a planet's nitrogen cycle.

Most importantly, nitrogen burial under completely abiotic and anoxic conditions on an Earth-like planet within the habitable zone is likely to be trivial compared to mantle degassing; hence an uninhabited Earth-like planet with a significant nitrogen inventory is unlikely to ever show low atmospheric N₂ pressures (Fig. 3a). This conclusion may be violated in a few cases, as follows:

(a) On young, very hot (>1000 K), reducing planets, N₂ may be rapidly reduced to NH₃ and dissolved in a magma ocean (Wordsworth, 2016). This scenario could probably be ruled out by inferring the planetary temperature through measurements of infrared emission, examination of the planet's atmospheric scale height to determine H₂ abundance, and/or observations of the host star to provide an estimate of the planet's age.

(b) Atmospheric pN₂ may be permanently low on planets that have lost their atmosphere by erosion and where the outgassing rate is at least an order of magnitude lower so that the atmosphere cannot be replaced (e.g., modern Mars). In this case, however, the abundance of other atmospheric gases would also be very low, and the planet's propensity to lose its atmosphere could be inferred from direct or indirect measurements of its mass and radius and therefore its surface gravity.

(c) Nitrogen burial could be more effective if abiotic N₂ fixation by volcanism, lightning, or impacts (Kasting and Walker, 1981; Kasting, 1990; Navarro-González et al., 1998; Mather et al., 2004) is at least an order of magnitude higher than estimated for early Earth. If the pH of the ocean on such a planet is significantly higher than 5, then even larger fixation rates would be required, because otherwise fixed NH₄ ⁺ (produced after conversion of NO₃ ^- to NH₄ ⁺ via hydrothermal reduction) (Brandes et al., 1998) would be returned to the atmosphere as NH₃ gas instead of adsorbing onto mineral surfaces (Stüeken, 2016). NH₃ gas quickly dissociates back to N₂ under UV light (Kuhn and Atreya, 1979). So far, a theoretical basis for unusually high extraterrestrial lightning rates is lacking. Enhanced volcanic activity may be detectable remotely through time-dependent observations of sulfate aerosols through transmission spectroscopy (Misra et al., 2015).

(d) Planets may have had a large compositional deficit of nitrogen after the initial period of accretion and enhanced atmospheric erosion by stellar UV light (Lichtenegger et al., 2010; Wordsworth and Pierrehumbert, 2014). This scenario may be detectable through the abundance of other volatiles in the planet's atmosphere or measurements of the nitrogen abundance in the host star (e.g., Brewer et al., 2016).

(e) Planets with a markedly lower oxygen fugacity in their mantle compared to that of Earth may not degas N₂, because mantle nitrogen may be stable as NH₃ and thus less volatile (Mikhail and Sverjensky, 2014; Li et al., 2015). But such planets may be discernable by the presence of CO rather than CO₂ in their atmospheres.

For planets that do not fall within the habitable zone, and thus are not covered by our model results, other scenarios could apply. For example, planets that are closer to the host star than the habitable zone that lack a surface ocean, such as Venus, would show insignificant nitrogen burial; hence atmospheric N₂ would increase as the mantle degasses. This effect would be enhanced on planets with a high oxygen fugacity where N₂ outgassing is favored over NH₃ retention, as proposed for early Venus (Wordsworth, 2016). Planets far beyond the habitable zone may have low atmospheric pN₂ if temperatures drop low enough for N₂ to become liquid or solid, such as on modern Pluto and possibly ancient Titan (McKay et al., 1993; Lorenz et al., 1997). Hence, pN₂ can fluctuate abiotically in such extreme cases, but for Earth-like planets within the habitable zone as considered in our model, abiotic N₂ drawdown is much less likely.

A biosphere on a completely anoxic Earth-like planet can potentially have substantial effects on atmospheric N₂ (Fig. 3b). A nitrogen burial flux equivalent to a few percent of modern biological N₂ fixation rates (without oxidative remineralization) may be sufficient to deplete the atmosphere of N₂ if mantle outgassing rates are comparable to those of Earth. In the absence of oxidative weathering, the only steady return fluxes of buried nitrogen back to the atmosphere would be metamorphism, volcanism, and mantle outgassing, and possibly catastrophic events (Section 3.3.2). It is important to note, however, that burial rates may be significantly different on planets that lack a surface reservoir equivalent to continental crust on Earth (Fig. 3d), which is able to take up and release atmospheric N₂ on hundred-million-year timescales.

Overall, our results strengthen the conclusion that the simultaneous presence of significant amounts of both N₂ and O₂ may be a biosignature and indicative of a biosphere with oxygenic photosynthesis (Schwieterman et al., 2015b; Krissansen-Totton et al., 2016) (Fig. 4). As shown above, a large anaerobic biosphere that never “invents” oxygenic photosynthesis can draw down N₂ to relatively low levels. Hence, both O₂ and N₂ would be low. Atmospheric erosion and the possibility of an unusually low mantle fugacity can be evaluated independently in such a scenario (e.g., Mars). On a completely abiotic planet orbiting in the habitable zone of a Sun-like star, O₂ can build up abiotically, but probably only under the condition that non-condensible gases (including N₂) are present in low amounts (Wordsworth and Pierrehumbert, 2014). According to this model, water from a surface ocean would be able to enter the upper atmosphere, where it would be photolyzed by UV, causing the H to escape and O to build up after surface sinks for oxidants are depleted. In this case, N₂ must start and remain low; otherwise the process is halted. Hence, high levels of abiotic O₂ would not coexist with a thick N₂ atmosphere. [Though note the likelihood that abiotic O₂ may be substantially higher for planets orbiting M dwarf stars due to other mechanisms not applicable to solar-type host stars (e.g., Harman et al., 2015; Luger and Barnes, 2015)]. A high-O₂-low-N₂ scenario is difficult to create biologically given the strong tendency of oxidative weathering and increasing oxygen fugacity in the mantle to release N₂ to the atmosphere. An inhabited planet whose biosphere invents oxygenic photosynthesis could eventually transition to oxidative weathering, thereby initiating rapid recycling of buried nitrogen from continental crust as on early Earth (e.g., Fig. 2c). This is thus perhaps the only case where both N₂ and O₂ reach high relative abundances in the atmosphere. In summary, (1) a planet with high pN₂ and no O₂ probably has either no biosphere (e.g., Venus) or a very small and/or young biosphere (e.g., first life on the Hadean Earth) that is incapable of transferring large quantities of nitrogen to the crust; (2) a planet with O₂ but no N₂ may be uninhabited; (3) a planet with neither O₂ nor high (modern) abundances of N₂ may host an anaerobic biosphere as exemplified by Archean Earth, provided that atmospheric erosion can be ruled out (cf. Mars); and (4) a planet with both significant N₂ and O₂ suggests the presence of a biosphere powered at least in part by oxygenic photosynthesis as on modern Earth. Low total N₂ on an anoxic planet (case 3) may be a weak biosignature, which could be confirmed through the detection of other biosignature gases or surface features (e.g., Des Marais et al., 2002; Pilcher, 2003; Domagal-Goldman et al., 2011; Schwieterman et al., 2015a).

FIG. 4.

Plausible interpretations of observed N₂ and O₂ abundances in exoplanetary atmospheres. Quantitative constraints on cutoffs for “low” and “high” abundances would need to be evaluated based on measurements of other gases and comparisons to other planets in the observed system.

5. Conclusion

The wide range of uncertainties in all our models, in particular about anything related to possible reconfigurations of Earth's mantle (Mikhail and Sverjensky, 2014), prohibits firm conclusions. Nevertheless, our results allow us to isolate a few key parameters for the evolution of a planet's nitrogen cycle and to formulate hypotheses about the evolution of atmospheric N₂ reservoirs on Earth and other planets. Some of these hypotheses may be testable with more constraints on nitrogen fluxes and with additional measurements of geological proxies for atmospheric pressure (Som et al., 2012; Marty et al., 2013; Glotzbach and Brandes, 2014; Kite et al., 2014; Som et al., 2016) (see Kavanagh and Goldblatt, 2015, for possible complications).

To first order, our results suggest that the greatest variability in atmospheric pN₂ over the history of a planet can be achieved if the planet is inhabited, if biomass burial is highly variable, and/or if it experiences oxygenation events or large-scale crustal melting. Other abiotic scenarios could be envisioned that could potentially lead to pN₂ fluctuations, such as N₂ freeze-out or atmospheric loss, but many of these cases would likely be discernable through other observations, in particular the orbit of the planet and the abundances of other gases. On an inhabited planet, variation in biomass burial can result from changes in the supply of metabolic substrates, including CO₂ (as on Earth, Fig. 2c, 2d) and N₂ (as in potential exoplanets, Fig. 3b). In the case of Earth, enhanced biomass burial in the Archean, followed by a stepwise shortening of the crustal residence time across the Paleoproterozoic and Neoproterozoic increases in oxidative weathering, could explain the drawdown and recovery of atmospheric N₂ inferred from abundances of N₂ in fluid inclusions at 3.5 Ga (Marty et al., 2013) and the size distribution of basaltic amygdales at 2.7 Ga (Som et al., 2016). We note that there is no independent evidence of enhanced burial of both organic carbon and carbonate in the Archean, because total organic carbon (TOC) and carbonate contents of Archean sedimentary rocks are not known to be particularly high. This discrepancy may suggest that large amounts of carbon-rich sediments and carbonated basalts have been subducted and lost. If one were to reject our hypothesis of enhanced nitrogen drawdown into continental crust as a temporary reservoir, then another alternative possibility for explaining a low Neoarchean N₂ pressure (Som et al., 2016) would be a much lower initial pN₂ value in the early Archean, followed by a marked increase in mantle degassing at some time between the Neoarchean and the modern. A test with our “F_org” model suggests that such a trajectory could be achieved if the mantle degassing rate were an order of magnitude higher throughout Earth's history than thought (Busigny et al., 2011). However, this scenario would be inconsistent with proposed N₂ pressures of 0.5–1.1 bar at 3.5 Ga (Marty et al., 2013). There is also no evidence for a major transition in the style or rate of mantle outgassing. But if such a transition occurred, it could conceivably be related to proposed changes in the mantle's redox state (Mikhail and Sverjensky, 2014; Nicklas et al., 2015; Aulbach and Stagno, 2016). Further research is needed to evaluate this possibility.

Lastly, our results may have some practical implications for observations of extrasolar planets. Despite all the uncertainties in our models, our results suggest that an anaerobic biosphere can—under Earth-like geological conditions—remove significant amounts of N₂ from the atmosphere. If multiple terrestrial planets around another star started out with similar volatile contents, but one of them has a significantly lower atmospheric N₂ abundance, then this may potentially serve as a biosignature. Measurements of other gases may be necessary to rule out atmospheric erosion as on Mars. In contrast, a planet with an oxygenic biosphere that stimulates oxidative weathering could maintain an atmosphere rich in both N₂ and O₂, similar to post-Archean Earth. Our results thus support the idea that the combination of N₂ and O₂ in an exoplanetary atmosphere may be a signature of a biosphere that is capable of oxygenic photosynthesis (Wordsworth and Pierrehumbert, 2014; Schwieterman et al., 2015b; Krissansen-Totton et al., 2016).

Footnotes

Acknowledgments

We thank the NASA postdoctoral program (E.E.S., E.W.S.), the NSF Graduate Research Fellowship program (M.A.K.), the NSF FESD program (grant number 1338810, subcontract to R.B.), the NSERC Discovery program (B.J.), the NASA Exobiology program (grant NNX16AI37G to R.B.) and the NAI Virtual Planetary Laboratory at the University of Washington (solicitation NNH12ZDA002C and Cooperative Agreement Number NNA13AA93A; E.W.S., R.B.) for financial support. We thank Jim Kasting and two anonymous referees for numerous helpful comments that improved the manuscript.

Author Disclosure Statement

No competing financial interests exist.

Abbreviations Used

References

Abbott

D.H.

and Isley

A.E.

(2002) The intensity, occurrence, and duration of superplume events and eras over geological time. J Geodyn, 34:265–307.

Ader

, Sansjofre

, Halverson

G.P.

, Busigny

, Trindade

R.I.

, Kunzmann

, and Nogueira

A.C.

(2014) Ocean redox structure across the Late Neoproterozoic oxygenation event: a nitrogen isotope perspective. Earth Planet Sci Lett, 396:1–13.

Ader

, Thomazo

, Sansjofre

, Busigny

, Papineau

, Laffont

, Cartigny

, and Halverson

G.P.

(2016) Interpretation of the nitrogen isotopic composition of Precambrian sedimentary rocks: assumptions and perspectives. Chem Geol, 429:93–110.

Algeo

T.J.

, Meyers

P.A.

, Robinson

R.S.

, Rowe

, and Jiang

G.Q.

(2014) Icehouse-greenhouse variations in marine denitrification. Biogeosciences, 11:1273–1295.

Arney

, Domagal-Goldman

S.D.

, Meadows

V.S.

, Wolf

E.T.

, Schwieterman

, Charnay

, Claire

, Hébrard

, and Trainer

M.G.

(2016) The pale orange dot: the spectrum and habitability of hazy Archean Earth. Astrobiology, 16, doi:10.1089/ast.2015.1422.

Aulbach

and Stagno

(2016) Evidence for a reducing Archean ambient mantle and its effects on the carbon cycle. Geology, 44:751–754.

Bahcall

J.N.

, Pinsonneault

M.H.

, and Basu

(2001) Solar models: current epoch and time dependences, neutrinos, and helioseismological properties. Astrophys J, 555:990–1012.

Berner

R.A.

(1982) Burial of organic carbon and pyrite sulfur in the modern ocean: its geochemical and environmental significance. Am J Sci, 282:451–473.

Berner

R.A.

(2006) Geological nitrogen cycle and atmospheric N₂ over Phanerozoic time. Geology, 34:413–415.

10.

Bjerrum

C.J.

and Canfield

D.E.

(2004) New insights into the burial history of organic carbon on the early Earth. Geochem Geophys Geosyst, 5, doi:10.1029/2004GC000713.

11.

Blake

R.E.

, Chang

S.J.

, and Lepland

(2010) Phosphate oxygen isotopic evidence for a temperate and biologically active Archaean ocean. Nature, 464:1029–1032.

12.

Boatman

C.D.

and Murray

J.W.

(1982) Modeling exchangeable NH₄ ⁺ adsorption in marine sediments: process and controls of adsorption. Limnol Oceanogr, 27:99–110.

13.

Bolton

E.W.

, Berner

R.A.

, and Petsch

S.T.

(2006) The weathering of sedimentary organic matter as a control on atmospheric O₂: II. Theoretical modeling. Am J Sci, 306:575–615.

14.

Borucki

W.J.

and Chameides

W.L.

(1984) Lightning: estimates of the rates of energy dissipation and nitrogen fixation. Rev Geophys, 22:363–372.

15.

Boudreau

B.P.

and Canfield

D.E.

(1988) A provisional diagenetic model for pH in anoxic porewaters: application to the FOAM site. Journal of Marine Research, 46:429–455.

16.

Brandes

J.A.

, Boctor

N.Z.

, Cody

G.D.

, Cooper

B.A.

, Hazen

R.M.

, and Yoder

H.S.

Jr. (1998) Abiotic nitrogen reduction on the early Earth. Nature, 395:365–367.

17.

Brewer

J.M.

, Fischer

D.A.

, Valenti

J.A.

, and Piskunov

(2016) Spectral properties of cool stars: extended abundance analysis of 1617 planet search stars. Astrophys J Suppl Ser, 225, doi:10.3847/0067-0049/225/2/32.

18.

Brewer

P.G.

and Murray

J.W.

(1973) Carbon, nitrogen and phosphorus in the Black Sea. Deep Sea Research and Oceanographic Abstracts, 20:803–818.

19.

Busigny

, Cartigny

, and Philippot

(2011) Nitrogen isotopes in ophiolitic metagabbros: a re-evaluation of modern nitrogen fluxes in subduction zones and implication for the early Earth atmosphere. Geochim Cosmochim Acta, 75:7502–7521.

20.

Canfield

D.E.

(2004) The evolution of the Earth surface sulfur reservoir. Am J Sci, 304:839–861.

21.

Cartigny

and Marty

(2013) Nitrogen isotopes and mantle geodynamics: the emergence of life and the atmosphere–crust–mantle connection. Elements, 9:359–366.

22.

Chang

and Berner

R.A.

(1999) Coal weathering and the geochemical carbon cycle. Geochim Cosmochim Acta, 63:3301–3310.

23.

Condie

and Korenaga

(2010) Thermal history of the Earth and its petrological expression. Earth Planet Sci Lett, 292:79–88.

24.

de Wit

M.J.

and Furnes

(2016) 3.5-Ga hydrothermal fields and diamictites in the Barberton Greenstone Belt—Paleoarchean crust in cold environments. Sci Adv, 2, doi:10.1126/sciadv.1500368.

25.

Des Marais

D.J.

, Harwit

M.O.

, Jucks

K.W.

, Kasting

J.F.

, Lin

D.N.C.

, Lunine

J.I.

, Schneider

, Seager

, Traub

W.A.

, and Woolf

N.J.

(2002) Remote sensing of planetary properties and biosignatures on extrasolar terrestrial planets. Astrobiology, 2:153–181.

26.

Dessler

A.E.

, Schoeberl

M.R.

, Wang

, Davis

S.M.

, and Rosenlof

K.H.

(2013) Stratospheric water vapor feedback. Proc Natl Acad Sci USA, 110:18087–18091.

27.

Domagal-Goldman

S.D.

, Meadows

V.S.

, Claire

M.W.

, and Kasting

J.F.

(2011) Using biogenic sulfur gases as remotely detectable biosignatures on anoxic planets. Astrobiology, 11:419–441.

28.

Driese

S.G.

, Jirsa

M.A.

, Ren

, Brantley

S.L.

, Sheldon

N.D.

, Parker

, and Schmitz

(2011) Neoarchean paleoweathering of tonalite and metabasalt: implications for reconstructions of 2.69 Ga early terrestrial ecosystems and paleoatmospheric chemistry. Precambrian Res, 189:1–17.

29.

Galloway

J.N.

, Dentener

F.J.

, Capone

D.G.

, Boyer

E.W.

, Howarth

R.W.

, Seitzinger

S.P.

, Asner

G.P.

, Cleveland

C.C.

, Green

P.A.

, Holland

E.A.

, and Karl

D.M.

(2004) Nitrogen cycles: past, present, and future. Biogeochemistry, 70:153–226.

30.

Garvin

, Buick

, Anbar

A.D.

, Arnold

G.L.

, and Kaufman

A.J.

(2009) Isotopic evidence for an aerobic nitrogen cycle in the latest Archean. Science, 323:1045–1048.

31.

Glotzbach

and Brandes

(2014) Air density of the Permian atmosphere: constraints from lithified raindrop imprints. Palaeogeogr Palaeoclimatol Palaeoecol, 409:280–289.

32.

Godfrey

L.V.

and Falkowski

P.G.

(2009) The cycling and redox state of nitrogen in the Archaean ocean. Nat Geosci, 2:725–729.

33.

Godfrey

L.V.

and Glass

J.B.

(2011) The geochemical record of the ancient nitrogen cycle, nitrogen isotopes, and metal cofactors. Methods Enzymol, 486:483–506.

34.

Goldblatt

, Claire

M.W.

, Lenton

T.M.

, Matthews

A.J.

, Watson

A.J.

, and Zahnle

K.J.

(2009) Nitrogen-enhanced greenhouse warming on early Earth. Nat Geosci, 2:891–896.

35.

Govindaraju

(1994) Compilation of working values and sample description for 383 geostandards. Geostandards Newsletter, 18:1–158.

36.

Greenfield

L.G.

(1991) Fixed ammonium in Antarctic rocks and soils and a possible cause of underestimation. Soil Biol Biochem, 23:397–399.

37.

Harman

C.E.

, Schwieterman

E.W.

, Schottelkotte

J.C.

, and Kasting

J.F.

(2015) Abiotic O₂ levels on planets around F, G, K, and M stars: possible false positives for life?. Astrophys J, 812, doi:10.1088/0004-637X/812/2/137.

38.

Hoch

M.P.

, Fogel

M.L.

, and Kirchman

D.L.

(1992) Isotope fractionation associated with ammonium uptake by a marine bacterium. Limnol Oceanogr, 37:1447–1459.

39.

Holland

H.D.

(2009) Why the atmosphere became oxygenated: a proposal. Geochim Cosmochim Acta, 73:5241–5255.

40.

Honma

(1996) High ammonium contents in the 3800 Ma Isua supracrustal rocks, central West Greenland. Geochim Cosmochim Acta, 60:2173–2178.

41.

Hren

M.T.

, Tice

M.M.

, and Chamberlain

C.P.

(2009) Oxygen and hydrogen isotope evidence for a temperate climate 3.42 billion years ago. Nature, 462:205–208.

42.

Huang

, Zhang

, Xia

, Hu

, and Son

S.W.

(2016) Is there a stratospheric radiative feedback in global warming simulations?. Clim Dyn, 46:177–186.

43.

Jia

, Kerrich

, Gupta

A.K.

, and Fyfe

W.S.

(2003) ¹⁵N-enriched Gondwana lamproites, eastern India: crustal N in the mantle source. Earth Planet Sci Lett, 215:43–56.

44.

Johnson

and Goldblatt

(2015) The nitrogen budget of Earth. Earth-Science Reviews, 148:150–173.

45.

Johnson

B.C.

and Melosh

H.J.

(2012) Impact spherules as a record of an ancient heavy bombardment of Earth. Nature, 485:75–77.

46.

Kadik

A.A.

, Kurovskaya

N.A.

, Ignat'ev

Y.A.

, Kononkova

N.N.

, Koltashev

V.V.

, and Plotnichenko

V.G.

(2011) Influence of oxygen fugacity on the solubility of nitrogen, carbon, and hydrogen in FeO-Na₂O-SiO₂-Al₂O₃ melts in equilibrium with metallic iron at 1.5 GPa and 1400 C. Geochemistry International, 49:429–438.

47.

Kanzaki

and Murakami

(2015) Estimates of atmospheric CO₂ in the Neoarchean–Paleoproterozoic from paleosols. Geochim Cosmochim Acta, 159:190–219.

48.

Kasting

J.F.

(1990) Bolide impacts and the oxidation state of carbon in the Earth's early atmosphere. Orig Life Evol Biosph, 20:199–231.

49.

Kasting

J.F.

(2013) What caused the rise of atmospheric O₂?. Chem Geol, 362:13–25.

50.

Kasting

J.F.

and Walker

J.C.

(1981) Limits on oxygen concentration in the prebiological atmosphere and the rate of abiotic fixation of nitrogen. J Geophys Res: Oceans, 86:1147–1158.

51.

Kasting

J.F.

, Pollack

J.B.

, and Ackerman

T.P.

(1984) Response of Earth's atmosphere to increases in solar flux and implications for loss of water from Venus. Icarus, 57:335–355.

52.

Kasting

J.F.

, Whitmire

D.P.

, and Reynolds

R.T.

(1993) Habitable zones around main sequence stars. Icarus, 101:108–128.

53.

Kavanagh

and Goldblatt

(2015) Using raindrops to constrain past atmospheric density. Earth Planet Sci Lett, 413:51–58.

54.

Kite

E.S.

, Williams

J.P.

, Lucas

, and Aharonson

(2014) Low palaeopressure of the martian atmosphere estimated from the size distribution of ancient craters. Nat Geosci, 7:335–339.

55.

Klingler

J.M.

, Mancinelli

R.L.

, and White

M.R.

(1989) Biological nitrogen fixation under primordial martian partial pressures of dinitrogen. Adv Space Res, 9:173–176.

56.

Koehler

M.C.

, Stüeken

E.E.

, Kipp

M.A.

, Buick

, and Knoll

A.H.

(2016) Spatial and temporal trends in Precambrian nitrogen cycling: a Mesoproterozoic offshore nitrate minimum. Geochim Cosmochim Acta doi:10.1016/j.gca.2016.10.050.

57.

Kopparapu

R.K.

, Ramirez

, Kasting

J.F.

, Eymet

, Robinson

T.D.

, Mahadevan

, Terrien

R.C.

, Domagal-Goldman

, Meadows

, and Deshpande

(2013) Habitable zones around main-sequence stars: new estimates. Astrophys J, 765, doi:10.1088/0004-637X/765/2/131.

58.

Krissansen-Totton

, Buick

, and Catling

D.C.

(2015) A statistical analysis of the carbon isotope record from the Archean to Phanerozoic and implications for the rise of oxygen. Am J Sci, 315:275–316.

59.

Krissansen-Totton

, Bergsman

D.S.

, and Catling

D.C.

(2016) On detecting biospheres from chemical thermodynamic disequilibrium in planetary atmospheres. Astrobiology, 16:39–67.

60.

Kuhn

W.R.

and Atreya

S.K.

(1979) Ammonia photolysis and the greenhouse effect in the primordial atmosphere of the Earth. Icarus, 37:207–213.

61.

Lee

C.T.A.

, Yeung

L.Y.

, McKenzie

N.R.

, Yokoyama

, Ozaki

, and Lenardic

(2016) Two-step rise of atmospheric oxygen linked to the growth of continents. Nat Geosci, 9:417–424.

62.

, Huang

, Wiedenbeck

, and Keppler

(2015) Nitrogen distribution between aqueous fluids and silicate melts. Earth Planet Sci Lett, 411:218–228.

63.

Libourel

, Marty

, and Humbert

(2003) Nitrogen solubility in basaltic melt. Part I. Effect of oxygen fugacity. Geochim Cosmochim Acta, 67:4123–4135.

64.

Lichtenegger

H.I.M.

, Lammer

, Grießmeier

J.M.

, Kulikov

Y.N.

, von Paris

, Hausleitner

, Krauss

, and Rauer

(2010) Aeronomical evidence for higher CO₂ levels during Earth's Hadean epoch. Icarus, 210:1–7.

65.

Lorenz

R.D.

, McKay

C.P.

, and Lunine

J.I.

(1997) Photochemically driven collapse of Titan's atmosphere. Science, 275:642–644.

66.

Luger

and Barnes

(2015) Extreme water loss and abiotic O₂ buildup on planets throughout the habitable zones of M dwarfs. Astrobiology, 15:119–143.

67.

Lyons

T.W.

, Reinhard

C.T.

, and Planavsky

N.J.

(2014) The rise of oxygen in Earth's early ocean and atmosphere. Nature, 506:307–315.

68.

Manabe

and Wetherald

(1967) Thermal equilibrium of the atmosphere with a given distribution of relative humidity. Journal of Atmospheric Sciences, 24:241–259.

69.

Manning

C.V.

, Zahnle

K.J.

, and McKay

C.P.

(2009) Impact processing of nitrogen on early Mars. Icarus, 199:273–285.

70.

Marty

(1995) Nitrogen content of the mantle inferred from N₂–Ar correlation in oceanic basalts. Nature, 377:326–329.

71.

Marty

and Tolstikhin

I.N.

(1998) CO₂ fluxes from mid-ocean ridges, arcs and plumes. Chem Geol, 145:233–248.

72.

Marty

and Zimmermann

(1999) Volatiles (He, C, N, Ar) in mid-ocean ridge basalts: assesment of shallow-level fractionation and characterization of source composition. Geochim Cosmochim Acta, 63:3619–3633.

73.

Marty

, Zimmermann

, Pujol

, Burgess

, and Philippot

(2013) Nitrogen isotopic composition and density of the Archean atmosphere. Science, 342:101–104.

74.

Mather

T.A.

, Pyle

D.M.

, and Allen

A.G.

(2004) Volcanic source for fixed nitrogen in the early Earth's atmosphere. Geology, 32:905–908.

75.

McKay

C.P.

, Pollack

J.B.

, Lunine

J.I.

, and Courtin

(1993) Coupled atmosphere-ocean models of Titan's past. Icarus, 102:88–98.

76.

Melosh

H.J.

and Vickery

A.M.

(1989) Impact erosion of the primordial atmosphere of Mars. Nature, 338:487–489.

77.

Mikhail

and Sverjensky

D.A.

(2014) Nitrogen speciation in upper mantle fluids and the origin of Earth's nitrogen-rich atmosphere. Nat Geosci, 7:816–819.

78.

Misra

, Krissansen-Totton

, Koehler

M.C.

, and Sholes

(2015) Transient sulfate aerosols as a signature of exoplanet volcanism. Astrobiology, 15:462–477.

79.

Nakamura

and Kato

(2004) Carbonatization of oceanic crust by the seafloor hydrothermal activity and its significance as a CO₂ sink in the Early Archean. Geochim Cosmochim Acta, 68:4595–4618.

80.

Navarro-González

, Molina

M.J.

, and Molina

L.T.

(1998) Nitrogen fixation by volcanic lightning in the early Earth. Geophys Res Lett, 25:3123–3126.

81.

Navarro-González

, McKay

C.P.

, and Mvondo

D.N.

(2001) A possible nitrogen crisis for Archaean life due to reduced nitrogen fixation by lightning. Nature, 412:61–64.

82.

Nicklas

R.W.

, Puchtel

I.S.

, and Ash

R.D.

(2015) The oxidation state of komatiites and the redox history of the mantle [abstract DI44A-07]. In AGU Fall Meeting Abstracts, American Geophysical Union, Washington, DC.

83.

Pennock

J.R.

, Velinsky

D.J.

, Ludlam

J.M.

, Sharp

J.H.

, and Fogel

M.L.

(1996) Isotopic fractionation of ammonium and nitrate during uptake by Skeletonema costatum: implications for δ¹⁵N dynamics under bloom conditions. Limnol Oceanogr, 41:451–459.

84.

Pierrehumbert

and Gaidos

(2011) Hydrogen greenhouse planets beyond the habitable zone. Astrophys J, 734, doi:10.1088/2041-8205/734/1/L13.

85.

Pilcher

C.B.

(2003) Biosignatures of early Earths. Astrobiology, 3:471–486.

86.

Rind

(1998) Just add water vapor. Science, 281:1152–1153.

87.

Rogers

L.A.

(2015) Most 1.6 Earth-radius planets are not rocky. Astrophys J, 801, doi:10.1088/0004-637X/801/1/41.

88.

Rosenfeld

J.K.

(1979) Ammonium adsorption in nearshore anoxic sediments. Limnol Oceanogr, 24:356–364.

89.

Roskosz

, Bouhifd

M.A.

, Jephcoat

A.P.

, Marty

, and Mysen

B.O.

(2013) Nitrogen solubility in molten metal and silicate at high pressure and temperature. Geochim Cosmochim Acta, 121:15–28.

90.

Schrag

D.P.

, Higgins

J.A.

, Macdonald

F.A.

, and Johnston

D.T.

(2013) Authigenic carbonate and the history of the global carbon cycle. Science, 339:540–543.

91.

Schwieterman

E.W.

, Cockell

C.S.

, and Meadows

V.S.

(2015a) Nonphotosynthetic pigments as potential biosignatures. Astrobiology, 15:341–361.

92.

Schwieterman

E.W.

, Robinson

T.D.

, Meadows

V.S.

, Misra

, and Domagal-Goldman

(2015b) Detecting and constraining N₂ abundances in planetary atmospheres using collisional pairs. Astrophys J, 810, doi:10.1088/004-637X/810/1/57.

93.

Seager

(2013) Exoplanet habitability. Science, 340:577–581.

94.

Sheldon

N.D.

(2006) Precambrian paleosols and atmospheric CO₂ levels. Precambrian Res, 147:148–155.

95.

Sloan

L.C.

and Pollard

(1998) Polar stratospheric clouds: a high latitude warming mechanism in an ancient greenhouse world. Geophys Res Lett, 25:3517–3520.

96.

Solomon

, Rosenlof

K.H.

, Portmann

R.W.

, Daniel

J.S.

, Davis

S.M.

, Sanford

T.J.

, and Plattner

G.K.

(2010) Contributions of stratospheric water vapor to decadal changes in the rate of global warming. Science, 327:1219–1223.

97.

Som

S.M.

, Catling

D.C.

, Harnmeijer

J.P.

, Polivka

P.M.

, and Buick

(2012) Air density 2.7 billion years ago limited to less than twice modern levels by fossil raindrop imprints. Nature, 484:359–362.

98.

Som

S.M.

, Buick

, Hagadorn

J.W.

, Blake

T.S.

, Perreault

J.M.

, Harnmeijer

J.P.

, and Catling

(2016) Earth's air pressure 2.7 billion years ago constrained to less than half of modern levels. Nat Geosci, 9:448–451.

99.

Stüeken

E.E.

(2013) A test of the nitrogen-limitation hypothesis for retarded eukaryote radiation: nitrogen isotopes across a Mesoproterozoic basinal profile. Geochim Cosmochim Acta, 120:121–139.

100.

Stüeken

E.E.

(2016) Nitrogen in ancient mud: a biosignature?. Astrobiology, 16:730–735.

101.

Stüeken

E.E.

, Buick

, Guy

B.M.

, and Koehler

M.C.

(2015) Isotopic evidence for biological nitrogen fixation by Mo-nitrogenase at 3.2 Gyr. Nature, 520:666–669.

102.

Stüeken

E.E.

, Kipp

M.A.

, Koehler

M.C.

, and Buick

(2016) The evolution of Earth's biogeochemical nitrogen cycle. Earth-Science Reviews, 160:220–239.

103.

Sundquist

E.T.

(1991) Steady-and non-steady-state carbonate-silicate controls on atmospheric CO₂ . Quat Sci Rev, 10:283–296.

104.

Sverdrup

H.U.

, Johnson

M.W.

, and Fleming

R.H.

(1942) The Oceans: Their Physics, Chemistry, and General Biology, Prentice Hall, New York.

105.

Walker

J.C.

, Hays

P.B.

, and Kasting

J.F.

(1981) A negative feedback mechanism for the long-term stabilization of Earth's surface temperature. J Geophys Res: Oceans, 86:9776–9782.

106.

Wasson

J.T.

and Kallemeyn

G.W.

(1988) Compositions of chondrites. Philos Trans A Math Phys Eng Sci, 325:535–544.

107.

Weiss

M.C.

, Sousa

F.L.

, Mrnjavac

, Neukirchen

, Roettger

, Nelson-Sathi

, and Martin

W.F.

(2016) The physiology and habitat of the last universal common ancestor. Nat Microbiol doi:10.1038/nmicrobiol.2016.116.

108.

Wetherill

G.W.

(1975) Late heavy bombardment of the Moon and terrestrial planets. Lunar and Planetary Science Conference Proceedings, 6:1539–1561.

109.

Wordsworth

and Pierrehumbert

(2014) Abiotic oxygen-dominated atmospheres on terrestrial habitable zone planets. Astrophys J Lett, 785, doi:10.1088/2041-8205/785/2/L20.

110.

Wordsworth

R.D.

(2016) Atmospheric nitrogen evolution on Earth and Venus. Earth Planet Sci Lett, 447:103–111.

111.

Young

G.M.

(1991) The geologic record of glaciation: relevance to the climatic history of Earth. Geoscience Canada, 18:100–108.

112.

Zahnle

K.J.

and Buick

(2016) Atmospheric Science: Ancient air caught by shooting stars. Nature, 533:184–186.

113.

Zahnle

K.J.

, Claire

M.W.

, and Catling

D.C.

(2006) The loss of mass-independent fractionation in sulfur due to a Palaeoproterozoic collapse of atmospheric methane. Geobiology, 4:271–283.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

1.15 MB

0.00 MB

Modeling p N 2 through Geological Time: Implications for Planetary Climates and Atmospheric Biosignatures

Abstract

1. Introduction

2. Model Setup

2.1. Biogeochemical nitrogen box model

2.2. Modeling effects on greenhouse gases and global climate

3. Results

3.1. Nitrogen burial constrained by carbon burial

3.1.1. Forg model

3.1.2. Forg + Heatflow model

3.1.3. Forg + Heatflow + Superplumes model

3.2. Atmospheric nitrogen evolution on abiotic and anoxic worlds

3.2.1. Abiotic anoxic planets

3.2.2. Biotic anoxic planets

3.3. Catastrophic events

3.3.1. Impacts

3.3.2. Superplumes

3.3.3. Planetary resurfacing

4. Discussion

4.1. The evolution of atmospheric pN2 on Earth

4.2. Climatic effects of atmospheric pN2 changes in the Archean

4.3. N2 in extraterrestrial atmospheres

5. Conclusion

Footnotes

Acknowledgments

Author Disclosure Statement

Abbreviations Used

References

Supplementary Material

3.1.1. F_org model

3.1.2. F_org + Heatflow model

3.1.3. F_org + Heatflow + Superplumes model

4.1. The evolution of atmospheric pN₂ on Earth

4.2. Climatic effects of atmospheric pN₂ changes in the Archean

4.3. N₂ in extraterrestrial atmospheres