The regulation of the air : a hypothesis

Introduction Conclusions References


Introduction
The modern air is a biological construction (Fig. 1).Air pressure is largely (78 %) determined by nitrogen.Nitrogen is a product of the biological nitrogen cycle, with N 2 made by denitrification and anammox reactions.It is fixed by nitrifying bacteria, with only small inorganic N fluxes via volcanic gases and lightning.The nitrogen cycle is closely inked to the carbon cycle (Falkowski, 1997).Nearly 21 % of the air is O 2 , which is the product of the extraction of carbon from CO 2 in the air by photosynthesis.CO 2 is now about 390 parts per million, but ∼280-300 ppm in the pre-industrial Holocene, and as low as 190 ppm in glacial maxima.Respiration, in which cells take up O 2 and release CO 2 , is the inverse side of the photosynthetic coin.On a planetary scale, the fluxes of photosynthesis and respiration are broadly reciprocal (e.g.Keeling and Shertz, 1992).Without photosynthetic oxygen, in a biosphere dominated by anaerobic metabolic processes such as fermentation, only limited cellular respiration could occur.The production of methane (now ∼1.8 ppm; ∼0.65 ppm in the Holocene, ∼0.35 ppm in glacial maxima) by methanogenesis is carried out by archaebactera using substrates such as acetate or hydrogen.Methanogenesis is largely dependent on the degradation of Introduction

Conclusions References
Tables Figures

Back Close
Full organic compounds and thus is in essence the recycling of photosynthetic productivity.
Even the biota living around hydrothermal vents on mid-ocean ridge volcanoes depend on the supply of sulphate from the oxygen-rich surface ocean (although methanogenesis can also occur using H 2 produced abiotically by the reaction between water and hot newly-erupted basalt).

Carbon transfers
The transfer of atmospheric CO 2 from the air to the interior of photosynthetic cells is expedited by a variety of mechanisms including the CO 2 concentrating mechanism (Badger et al., 2005).The delivery of CO 2 to photosynthetic organelles is not a trivial accomplishment.Despite this, carbon entry into the cell is generally efficient, with little fractionation between 12 C and 13 C.This high isotopic efficiency implies a high yield and thus nearly all the carbon accessed from the external environment is transferred into the cell.Once inside the cell the rubisco enzyme, the most abundant protein on Earth, catalyses the fixation of CO 2 .There are various forms of rubisco.Rubisco I, the most abundant, facilitates oxygenic photosynthesis.As an carboxylase/oxygenase, rubisco operates with either CO 2 or O 2 as substrate, either selecting CO 2 from the atmosphere to capture C into organic life, or acting in reverse.As the enzyme is highly selective, only a part of the accessible carbon dioxide inventory within the cell is chosen for incorporation into organic product.Thus, rubisco-catalysed carbon fixation causes strong isotopic fractionation.
Rubisco I may have evolved nearly 3 Ga ago (Nisbet et al., 2007), yet rubisco's "specificity" (its preference for CO 2 over O 2 ) remains paradoxically restricted.In contrast with other enzymes, rubisco seems inefficient, with a sluggish reaction rate (Tcherkez et al., 2006).However, rubisco's 'inefficiency' may be deceptive.On a molecular level, analysis of the kinetics and isotopic fractionation of Rubisco I in oxygenic photosynthetic organisms (Tcherkez et al., 2006)  to a near-perfect compromise between catalytic turnover and CO 2 : O 2 specificity (Gutteridge and Pierce, 2006).Organisms that that are adapted to higher CO 2 : O 2 ratios contain rubisco with higher catalytic rates and lower specificity.In contrast, a rubisco enzyme with a higher specificity (i.e. more able to select for CO 2 at low CO 2 : O 2 ratios) will operate only at lower temperatures and lower rates (Tcherkez et al., 2006), and will thus suit organisms adapted to a different range of habitats on the climatically varied planet.In each photosynthetic species, the rubisco enzymes may be nearly optimally fitted to their specific tasks, each exquisitely tuned to the local environment (Tcherkez et al., 2006;Gutteridge and Pierce, 2006), under a globally mixed atmosphere.On a macroscopic scale, the CO 2 "compensation point", Γ CO 2 , defines the equilibrium between CO 2 uptake by photosynthesis and release by photorespiration.In plants, the CO 2 compensation is the CO 2 concentration at which net CO 2 fixation is zero (e.g.Tolbert et al., 1995).As carbon dioxide re-supply from the external environment (either from the air or dissolved gases in seawater which in turn have equilibrated with the air) is generally not limiting, this suggests that the fundamental molecular control on the macroscopic compensation barrier, Γ CO 2 , lies in the biochemistry of rubisco catalysis.If the CO 2 content lies on the wrong side of the barrier limit in the external environment, net plant and algal growth ceases.If CO 2 is too low, organisms cannot capture carbon.O 2 and Γ CO 2 are interlinked, with linear co-variance (Tolbert et al., 1995, but see Andr é, 2011a, b also).Tolbert et al. (1995) postulated an O 2 compensation point, defined as the upper limit of O 2 at a given CO 2 level, above which plants cannot have positive carbon uptake.Their experimental data rather paradoxically implied two distinct functions, one linking Γ CO 2 to O 2 , and the other linking Γ O 2 to CO 2 , although theoretically there should be a single linear function.The experimental results of Tolbert et al. (1995) led Nisbet andNisbet (2008) to propose the existence of a "prohibited region" in O 2 : CO 2 space, where oxygen partial pressure would be so high that plant growth would stop, and eventually the plant would respire itself to death.showed that the Tolbert et al. (1995) experiments may have been affected by problems such as leakage artefacts.
Oxygen compensation Γ O 2 (stated in %) is related to rubisco specificity and CO 2 concentration, C. Crucially, Andr é (2011a) demonstrated that only a single reciprocal function exists linking Γ CO 2 and Γ O 2 .Andr é (2011a) defined a plant-specificity-factor (Sp) which took into account both the specificity of the rubisco enzyme and also the CO 2 gradient between the atmosphere and the Rubisco site.Where E is the gross Oxygen evolution, a parameter linked to the electron transport rate generated by light in chloroplasts, and R is the rate of respiration, Andr é (2011a) showed that a finite or realistic Γ O 2 cannot be reached in a closed biosystem without a high rate of respiration (R).
Andr é (2011a) showed that the CO 2 and O 2 exchanges of the green (productive) part of plants can be fitted, in large range of environmental parameters, by a simple model that mimics the Rubisco response to O 2 and CO 2 variations.In a closed biosystem, Andr é (2011a) showed that, dependent on plant specificity Sp, crossing points of oxygenation and carboxylation gas exchanges exist, between net photosynthesis and the opposite net photorespiratory processes, where net photosynthetic uptake of CO 2 equals gross O 2 uptake.Taking O x as the oxygen concentration (in %) at the crossing point, and C x as the CO 2 concentration (in ppm), the crossing point is expressed by reciprocal relationships (Andr é, 2011a): For average C3 plants, with an average plant specificity Sp of 0.132 % ppm −1 , Andr é's (2011a) relationship gives CO 2 of 317 ppm for the modern atmospheric mixing ratio of dioxygen of 20.95 %.CO 2 in the pre-industrial air had a mixing ratio of roughly Introduction

Conclusions References
Tables Figures

Back Close
Full This leads to the central hypothesis (Tolbert et al., 1995;Igamberdiev and Lea, 2006;Nisbet et al., 2007;Nisbet and Nisbet, 2008;Andr é, 2011a) that Rubisco controls the balance of CO 2 and O 2 in the air.Assuming a very simplified global system consisting only of C3 land plants (with a specificity of 0.132 % O 2 per ppm CO 2 ) and atmosphere (omitting oceans and the geological cycling of carbon dioxide in the ocean-atmosphere system), the fit is remarkable.As Andr é (2011a) points out, "the O 2 and CO 2 concentrations of the pre-industrial atmosphere are the compensation points of the global biosystem".
But is this accidental, or has it evolved over time?If "the present is the key to the past" the specificity of rubisco may have helped sustain planetary habitability over geological time.Nisbet and Nisbet (2008) considered the greenhouse implications of this hypothesis, suggesting that, by setting the O 2 : CO 2 ratio, the evolution of the enzymatic specificity of Rubisco I, as it controlled plant specificity, may have controlled the planetary greenhouse ever since 3Ga ago, and hence also the maximum abundance of CH 4 and water vapour in the air.Andr é (2011a), suggested the O 2 : CO 2 ratio is not merely fortuitous but is a consequence of a dynamic stabilisation of the atmosphere.More generally, Andr é (2011a) strongly supported the hypothesis of co-evolutionary plant-atmosphere-climate feedbacks on the modern planet.

The Greenhouse and its geological record
If the atmospheric O 2 : CO 2 ratio is biologically set, the obvious evolutionary hypothesis is that the planetary greenhouse, the Earth's surface temperature and the acidity of the oceans, has been biologically set by natural selection over the aeons (Nisbet, 2002;Nisbet and Nisbet, 2008).Evolutionary competition, by natural selection of rubisco, has produced a mechanism that inherently has the ability to set the greenhouse warming of Introduction

Conclusions References
Tables Figures

Back Close
Full the air.But did rubisco specificity co-evolve with the global greenhouse?Can natural selection sustain an equable global temperature?Or is the clemency of the planetary climate over the aeons simply an inorganic geochemical consequence of the nature of the planetary surface?Planetary surface temperature is governed by CO 2 , supplemented by CH 4 and minor gases, and indirectly by N 2 and O 2 pressure.The dominant greenhouse gas, water vapour in the air, can be thought of as a "responding" feedback to variations in gases such as CO 2 , CH 4 and N 2 O, rather than a "controlling" gas.The modern atmosphere is globally mixed over a period of a year, and equilibrates quickly with the ocean surface and upper waters, a process speeded by planktonic use of carbonic anhydrase.
Stability in the ocean/atmosphere system is conferred by the very large (and longlived) biologically-created reservoirs of atmospheric N 2 and O 2 , and reciprocally by the burial of reduced organic matter, which is recycled via volcanoes by plate tectonics.
In the oxygen-rich atmosphere created by photosynthesis, any emitted CH 4 and CO molecules are within a few years converted to CO 2 .Rates of biological processes vary exponentially with temperature.This means that, compared to geological timescales, biological feedback controls can in principle act rapidly, controlling CO 2 and CH 4 , and thereby managing the temperature of the ocean/atmosphere system.Inorganic processes have a longer time-scale.Atmospheric CO 2 , through acid rain, influences the rate of chemical erosion and the supply of alkalinity by rivers to seawater.
By equilibrating with ocean water, atmospheric CO 2 helps to set ocean pH.This in turn influences carbonate precipitation and thus the inorganic feedback controls on CO 2 (Walker et al., 1981; see also Fig. 8.2 in Kasting, 2008).Carbonate precipitation, and hence CO 2 in the air, also depends on the alkalinity of the ocean, set by the combined effects of river runoff (which is influenced by atmospheric CO 2 ) and mid-ocean ridge hydrothermal processes (which are less linked to the atmosphere but do depend in part on the CO 2 content of the seawater).As the Earth has evolved, the continental area may have changed over time (but see discussion in Arndt and Nisbet, 2011).Moreover, as the potential temperature of the mantle has cooled, thickening lithosphere may have Introduction

Conclusions References
Tables Figures

Back Close
Full became more capable of supporting mountains and hence more erosion of silicates, taking up CO 2 .Thus there may be purely inorganic feedbacks capable of changing and managing the greenhouse over time as the external inputs varied.
4 Long-term controls on the planetary atmosphere and temperature Over time, the Sun has steadily warmed (Sagan and Chyba, 1997;Goldblatt et al., 2009;Nisbet and Nisbet, 2008;Ueno et al., 2009;Rosing et al., 2010; see also Fig. 8.1 in Kasting, 2008).It might be expected that, under the faint young Sun, the Hadean and Archaean Earth would have been covered in ice (see Nisbet and Sleep, 2001;Nisbet and Fowler, 2004).If the oceans are cool now, in a modern world subject to repeated ice ages, then billions of years ago under a faint Sun the planet ought to have existed in deep permafrost, too cold for life except near volcanic vents.However, the evidence in Jack Hills zircons in Western Australia that implies the presence of liquid water oceans 4.3 Ga ago (Wilde et al., 2001), and the presence in the 3.8 Ga old Isua belt of sedimentary rocks (Rosing, 1999) tells us that surface temperatures permitted liquid water.High concentrations of atmospheric greenhouse gases, either CO 2 or perhaps methane, could have prevented the oceans from freezing over.However, from the presence of magnetite and the absence of siderite in Archaean sediments, Rosing et al. (2010) infer that CO 2 levels in the Archean atmosphere were little greater than those of today.Nevertheless, in the mid-Archaean, around 3.2 Ga ago, Hessler at al. (2004) showed that the minimum partial pressure of carbon dioxide was several times higher than present-day values.Grotzinger and Kasting (1993) showed that even at CO 2 partial pressures of 0.1 bar, the pH of the ocean need not have been much different from the modern value of 8.1.
It appears that the planetary thermostat has been adjusted down as the Sun has warmed up.A wholly inorganic control is plausible (Holland, 1984), initially with high atmospheric CO in the early Precambrian, and then with lower CO 2 as the Earth aged and the Sun warmed, while the continental crust evolved, the lithosphere thickened and Introduction

Conclusions References
Tables Figures

Back Close
Full mountains grew taller.The anthropic principle is relevant here: we may live on a planet whose inbuilt inorganic geochemical evolution, via changes in the composition and area of the surface, and the erodibility of continental crust, coupled with the cooling of the mantle, has over the aeons matched the warming of the Sun with a declining atmospheric greenhouse.Is the long-term maintenance of life-sustaining conditions on Earth pre-ordained, with the evolving biosphere a passive occupant of the inorganic aging of the planetary environment?Or can the physical environment itself be malleable, moulded by evolution, so that the physical setting itself is reshaped as the biosphere evolves?Can biology, by its command of kinetics overrule the long-term equilibria of geological supplies and sinks between the Earth's interior and surface?Can rapid biological processes, with their large carbon fluxes, overwhelm the smaller fluxes of the slower inorganic controls?By sequestering reservoirs such as atmospheric oxygen and buried sedimentary carbon, and by setting parameters such as temperature and pH, can biological processes control thermodynamic equilibria on a planetary scale?
There is strong controversy over the temperature of the oceans in the mid-Archaean.From isotopic evidence in cherts, Knauth and Lowe (2003), and Robert and Chaussidon (2006) suggested temperatures up to 50-70 • C. A methane greenhouse would be the obvious explanation if temperatures really were so high.Almost all atmospheric methane is a biological product, and there is good evidence for mid-Archean methanogenesis (e.g.Grassineau et al., 2006), so a methane-rich atmosphere is indeed possible.If so, the biological cycle of methanogenesis, balanced by methanotrophy, atmospheric and sedimentary oxidation (including anaerobic oxidation) was controlling the mid-Archaean global surface temperature.But did biological control continue?

Feedbacks in rubisco evolution
About 2.9 Ga ago, large scale sedimentary calcite and dolomite deposits appear, deposited in shallow water (Nisbet et al., 2007).Moreover, since then except in a few Introduction

Conclusions References
Tables Figures

Back Close
Full notable but isolated time periods, the C isotope ratio in sedimentary carbonates has been broadly modern in aspect (Schidlowski, 1988).The implication is that the controls on carbon partition between organic matter and carbonate deposition into sediment, whether inorganic or biologically precipitated, have been broadly unchanged from 2.9 Ga ago to now.
Photosynthesis is today the key biological control on CO 2 (Hahn, 1987(Hahn, , 1991)), and also indirectly controls CH 4 production from dead organic matter.Hence on an annualto millennium-scale, the global temperature (set by the CO 2 and CH 4 greenhouse) must be determined by the primary control on CO 2 partitioning into organic matter, which is rubisco's specificity (Tcherkez et al., 2006;Gutteridge and Pierce, 2006).If rubisco were more effective at drawing down CO 2 from the air (e.g. with a take-over of the biosphere by C4 plants, so that the overall average plant specificity changed markedly), eventually sequestering organic carbon on the seabed, the water would freeze.If rubisco were less effective, the oceans would become acidic, hot, and deoxygenated.
Rubisco is a biological product and like all biology is subject to natural selection.As the environment changes, evolution will act on rubisco.It is thus a tenable hypothesis that, at least in part, the broad stability of planetary temperature over the aeons occurred because natural selection of competing rubisco molecules favoured those biological processes that, by managing the atmosphere, maximised the likelihood of the survival of microbial consortia.
Consider the plankton, both eukaryotic and prokaryotic.Intrinsically, only a few species would be expected, equal in number to the number of bio-limiting resources -N, P, Fe, Si, inorganic carbon, light energy, trace metals.Yet thousands of plankton species coexist in natural waters.This paradox is resolved by the hypothesis (Huisman and Weissing, 1999) that resource competition creates oscillating and chaotic populations.These oscillations allow many species to coexist, although depending on only a few discrete resource types.Natural selection, presumably including selection for rubisco specificity in photosynthesisers, operates on the oscillating populations, to select for survivability.Separately, there is evidence that large variations in key indicators such as the N:P ratio occur in the various microbial "biomes" in the ocean.While the bulk global ocean N:P ratio (16:1) is very highly conserved on a planet-wide basis, the local N:P can vary widely, from 11:1 to 20:1 (Weber and Deutsch, 2010).In the global ocean, reservoirs of N and P may be kept near the Redfield ratio (C:N:P = 106:16:1, the stoichiometric molecular ratio of carbon, nitrogen and phospohorus in oceanic plankton), by competition between species that add N to the ocean through biological N 2 fixation when N is limiting, but when it is not limiting are out-competed by other plankton (Weber and Deutsch, 2010).Collectively in the global ocean circulation, the biomes manage a very efficient global regulation of N:P, yet individual biomes can be very different, each selected to be best adapted for local conditions.If this is so for microbial N and P, could a similar evolutionary mechanism control CO 2 ?
In the ocean, microbial biomes based on photosynthetic plankton have climate impact.Selection over time will favour biomes that are best fitted to survive by producing favourable responses from the environment, even if those benefits are only marginal.
Over millions of years, selection would favour phytoplankton species which contained rubisco with specificities that maximised climate favourability.Ecological consortia that act adversely to climate will be selected against.In extreme cases, global over-heating or rare catastrophic glaciations would eliminate organisms whose blooms destabilised global climate.Over time, by maximising survival, repeated selective evolutionary feedbacks would fine-tune microbial rubisco specificities as the rubiscos in turn controlled the greenhouse impact of the atmosphere.
On land, similar selective pressures must apply.Rainforests substantially manage local and regional climates, controlling albedo via leaf colour and also by trapping and transpiring water and emitting aerosols to nucleate cloud cover.On a large scale, major forest biomes strongly influence global weather, for instance by influencing the location of major convergence zones and the jet streams.Similarly, in the boreal and tropical forests, carbon sequestration by peat bogs has global impact by drawing down CO 2 , a global cooling impact that is countered by methane emission.Introduction

Conclusions References
Tables Figures

Back Close
Full In the relatively recent geological past C4 plants have evolved (Sage, 2004).These plants, which through convergent evolution have appeared in many plant groups, preconcentrate CO 2 around the rubisco, and thus can thrive at lower levels of ambient CO 2 than C3 plants.Globally, they are common at times of ice ages, when atmospheric CO 2 may be below 200 ppm.In these circumstances, the pre-concentration stage clearly gives them advantage over C3 plants.Contrarily, higher CO 2 in interglacial periods gives advantage to C3 plants (Andr é, 2011b).Interestingly, C4 plants (especially grasses) often form fire ecologies: though fires immediately return C to the air, their long-term impact is to remove C by sequestering it as charcoal, which is stable over millions of years.The Holocene global CO 2 level (roughly 290 ppm) may have been set (Andr é, 2011b) by competition between the dominant C3 plants (favouring ∼320 ppm CO 2 at ∼21 % O 2 ) and the specialist C4 plants (favouring lower CO 2 values).
More generally, in the oxygen-rich atmosphere, on a warm planet at a time of increased atmospheric CO 2 (say from a volcanic injection), carbon fixation by photosynthesis will increase.In these circumstances, with CO 2 abundantly accessible, it is possible that natural selection will favour photosynthetic organisms that use a catalytically faster variant of rubisco.Eventually photosynthesis will use up the excess CO 2 , reducing greenhouse warming by drawing down CO 2 towards Γ CO 2 .In the cooler low-CO 2 environment that results, there will be a greater requirement for rubisco variants that have greater specificity for CO 2 over O 2 (Tcherkez et al., 2006), but this adaptation means draw-down rates will be reduced.Eventually, as organic matter is recycled, the ratio of CO 2 : O 2 will rebalance temperature again.The very large O 2 reservoir, which changes only slowly as it is much larger than the annual flux, serves as a stabiliser on the system, drawing CO 2 back to long-term average levels.occurs, such as a volcanic eruption or solar warming, then there will be a proportional response as photosynthesis and respiration increase/decrease or decrease/increase, taking up or releasing CO 2 .Methane too will change, especially as temperature change affects water precipitation in rain and snow.Secondly, there will also be an integrated response, as carbon is sequestered or released, for example by precipitation or dissolution of calcite, or by the deposition or release of organic matter, for example in sea bed organic deposits, or in peat bogs or fires.Moreover, there will be a strong integrated response from methane recycling organic matter.Thirdly, in addition to this, there will be a derivative response, as organisms are favoured or selected against, according to their balance between rubisco specificity and the rate of catalytic turnover and productivity.Over the longer term via this natural selection for specificity, especially if the external perturbation is prolonged (e.g.solar warming), the control will be adaptive, iteratively ramping up so that evolution will change the contemporaneous optimum temperature at which the biosphere has maximum productivity.For example, in future as the aging Sun brightens, dynamically adaptive evolution will favour C4 plants, capable of thriving in low-CO 2 air over C3.This will stabilise global temperature by reducing the atmospheric greenhouse as the solar input increases.Thus there is circumstantial but permissive evidence to support the hypothesis that rubisco manages the atmosphere and has done so for the past three billion years.

Testable implications of the hypothesis of co-evolution between rubisco and climate
This hypothesis implies testable propositions: 1.That inorganic controls on carbonate precipitation are not the primary controls on atmospheric CO 2 and O 2 , but rather that atmospheric abundances of CO 2 , O 2 and CH 4 are primarily set by biological control.This biological control determines the wider environment (pressure, temperature, acidity, redox) in which inorganic Introduction

Conclusions References
Tables Figures

Back Close
Full reactions occur.The proposition predicts that atmospheric mixing ratios of CO 2 , O 2 and CH 4 , and their variability, are fully explicable by biological processes.
2. That, in addition to management of carbon dioxide, biological sequestration of nitrogen into the atmosphere (which creates atmospheric air pressure) has overwhelmed the inorganic (geological) processes that otherwise would be in control.
This predicts that nitrogenase, often thought of as inefficient, must be as well tuned to its task as Rubisco I.
3. That for any planet, chance trajectories of surface temperature which sustain clement conditions over geological aeons are very improbable, given the complex parameters of solar brightening, geological partitioning of C, O, N and S species between the planetary interior and the surface, exchange between ocean and air; and time.In the Earth's early history, inorganic processes may have produced clement conditions for a relatively brief period (10 5 -10 7 yr) (Sleep et al., 2001), but this could not have been sustained over aeons.The implication is that it will be very difficult to create an inorganic feedback model that successfully sustains clement temperatures over the aeons.In contrast, the continually selective nature of biological evolution has inbuilt adaptation.
Discussion Paper | Discussion Paper | Discussion Paper | suggests that evolution has led Discussion Paper | Discussion Paper | Discussion Paper | However this was shown to be incorrect by Andr é (2011a, b), who re-investigated the high-O 2 prohibited zone hypothesis.He Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | atmospheric 0.03 % CO 2 and 21 % O 2 equilibria seem to be limits set by the average specificity properties of rubisco from plants and algae".
Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

Figure 3 ,
Figure 3, modelled on industrial PID (proportional-integral-derivative) controllers, illustrates the way natural selection, acting on rubisco specificity, can manage the surface temperature.The "optimum" temperature is the temperature at which the contemporaneous global biosphere has maximum productivity.If an external perturbation Discussion Paper | Discussion Paper | Discussion Paper |

Fig. 1 .Fig. 2 .Fig. 3 .
Fig. 1.Biological controls on atmospheric composition, pressure and temperature.Positive and negative temperature forcing is shown by + and − symbols.Green shading on land represents photosynthetic life.Abbreviations: (AOM) Anaerobic methane oxidation.AIR -natural mix, now mainly N 2 and O 2 but previously with CO 2 and CH 4 .Other letters are chemical species.