Reviews and syntheses: Heterotrophic fixation of inorganic carbon – significant but invisible flux in environmental carbon cycling

Heterotrophic CO2 fixation is a significant yet underappreciated CO2 flux in environmental carbon cycling. In contrast to photosynthesis and chemolithoautotrophy – the main recognized autotrophic CO2 fixation pathways – the importance of heterotrophic CO2 fixation remains enigmatic. All heterotrophs – from microorganisms to humans – take up CO2 and incorporate it into their biomass. Depending on the availability and quality of growth substrates, and drivers such as the CO2 partial pressure, heterotrophic CO2 fixation contributes at least 1 %–5 % and in the case of methanotrophs up to 50 % of the carbon biomass. Assuming a standing stock of global heterotrophic biomass of 47–85 Pg C, we roughly estimate that up to 5 Pg C might be derived from heterotrophic CO2 fixation, and up to 12 Pg C yr−1 originating from heterotrophic CO2 fixation is funneled into the global annual heterotrophic production of 34–245 Pg C yr−1. These first estimates on the importance of heterotrophic fixation of inorganic carbon indicate that this pathway should be incorporated in present and future carbon cycling budgets.


Introduction
Fixation of CO 2 is a fundamental biosynthetic process in nature (Beer et al., 2010;Berg et al., 2007), providing the main source of metabolic energy on Earth (Giovannoni and Stingl, 2005).At the same time, it acts as a sink for atmospheric CO 2 , the most important greenhouse gas, which is responsi-ble for more than 60 % of the "enhanced greenhouse effect" resulting in global warming (Beer et al., 2010;Berg, 2011;Houghton, 2007;Le Quéré et al., 2016).
While photosynthesis and chemosynthesis are the most important processes of carbon fixation, non-autotrophic carbon fixation, i.e., the carbon fixation mediated by heterotrophic organisms, might also be relevant, albeit uncommonly quantified.While heterotrophs are, per definition, organisms that respire organic compounds to gain energy and build up biomass, CO 2 fixation plays also an essential role in heterotrophic carbon metabolism.The diversity of carboxylating enzymes in nature reaches far beyond autotrophy and virtually all heterotrophs harbor numerous enzymes fixing dissolved inorganic carbon.Even though the first carboxylase in heterotrophs was discovered more than 80 years ago (Wood and Werkman, 1936), the role of heterotrophs in carbon cycling has so far largely focused on the oxidation of organic substrates using oxygen or alternative electron acceptors (e.g.nitrate, ferric iron, sulfate) and the production of CO 2 .Similar to the CO 2 fixation by autotrophs, "heterotrophic CO 2 fixation" might, however, constitute a significant carbon flux in specific habitats.The relevance of this process has hardly been quantified due to the lack of reliable estimates of heterotrophic CO 2 fixation for most organisms and habitats, and the presumption that CO 2 fixation in natural environments is restricted to autotrophic organisms.
To fill this gap, we review the current knowledge on (i) the role of heterotrophic CO 2 fixation for cellular Published by Copernicus Publications on behalf of the European Geosciences Union.metabolism, (ii) respiration and non-autotrophic CO 2 fixation, and (iii) CO 2 fixation in habitats dominated by heterotrophs, and we provide (iv) quantitative estimates of heterotrophic CO 2 fixation in different environments.
The most important CO 2 fixation pathway in all organisms is anaplerosis.Anaplerosis replenishes intermediates in the tricarboxylic acid (TCA) cycle, which have been released for biosynthesis.TCA metabolites are used as building blocks for macromolecular compounds, e.g.almost half of all amino acids in prokaryotes are directly synthesized from oxaloacetate and α-ketoglutarate (Fuchs, 1999).For this purpose, heterotrophs use the enzymes pyruvate carboxylase present in a large variety of organisms, including prokaryotes, archaea, yeasts, fungi, and higher organisms (e.g.mammals), and phosphoenol pyruvate (PEP) carboxylase, widely distributed in bacteria (Attwood, 1995;Jitrapakdee and Wallace, 1999;Sauer and Eikmanns, 2005;Jitrapakdee et al., 2008) (Fig. 1).The replenishment of metabolites continuously withdrawn from the TCA cycle via the anaplerotic reaction of PEP carboxylase entails an assimilation of CO 2 corresponding to 25 % of the initial substrate's carbon content.In a systematic stable isotope labeling experiment with Bacillus subtilis, a gram-positive heterotrophic bacterium widespread in the environment, the interdependency of pathways and rates of CO 2 fixation on the concurrent utilization of organic substrate(s) was explored (Spona-Friedl et al., 2020).Over the course of the experiments B. subtilis assimilated 6 % and 5 % of carbon biomass from the external H 13 CO 3 pool when growing on glucose and lactate, respectively (Spona-Friedl et al., 2020).Growth on malate, an intermediate of the TCA cycle, expected to serve directly to refill the oxaloacetate pool of the TCA cycle, still revealed a contribution to biomass production from inorganic carbon of 3 % (Spona-Friedl et al., 2020).In fact, PEP carboxylase was still actively transforming pyruvate to oxaloacetate.Heterotrophic CO 2 fixation continued to a lower extent even in the absence of cell growth during the stationary phase (Spona-Friedl et al., 2020), indicating that anaplerotic reactions are important in low-productivity habitats (see below).
Overall, heterotrophic CO 2 fixation via anaplerosis in microorganisms contributes around 1 % to 8 % to the carbon biomass (Romanenko, 1964;Perez and Matin, 1982;Doronina and Trotsenko, 1984;Miltner et al., 2004;Roslev et al., 2004;Hesselsoe et al., 2005;Šantrůčková et al., 2005;Feisthauer et al., 2008;Akinyede et al., 2020;Spona-Friedl et al., 2020).Under particular environmental conditions even higher contributions were reported (Perez and Matin, 1982).The advantage that CO 2 is readily available to the cell either as atmospheric gas or, more commonly, in its hydrated form HCO − 3 obviously outcompetes the disadvantage that carboxylation is generally an endergonic reaction (Faber et al., 2015).This thermodynamic obstacle may be less important when carboxylation supports the assimilation of organic substrates more reduced than the organism's biomass, resulting in carbon-limited but excess-energy conditions (Heijnen and Roels, 1981;Ensign et al., 1998;von Stockar et al., 2006;Battley, 2013).In this case, in addition to anaplerosis further carboxylation reactions are induced (Fig. 1) to add oxidized C (from CO 2 ) to the reduced organic substrate for adjusting the degree of reduction to that of the biomass (Fig. 2).For example, the assimilation of leucine and propionate into biomass entails carboxylation of the initial C-6 and C-3 carbon bodies, respectively, and thus triggers an assimilation of dissolved inorganic carbon (DIC) that corresponds to 17 % and 33 % of the initial substrate's carbon content, respectively (Erb, 2011).In aerobic methane oxidation, the full oxidation potential of one molecule of CO 2 is needed to adjust the high degree of reduction of methane to that of biomass during its assimilation.Consequently, methanotrophs derive up to 50 % of their carbon biomass from CO 2 (Strong et al., 2015;Battley, 2013).
Besides the degree of reduction of organic carbon sources, the partial pressure of CO 2 plays a role.Carboxylases may catalyze carboxylation as well as decarboxylation of organic compounds, and the equilibrium of the reaction depends on the concentrations of all compounds involved.An increase in the CO 2 concentration may move the equilibrium of the reaction toward the product of the carboxylation, and thus lead to an increase in CO 2 fixation.
In a physiological context, the amount of inorganic carbon fixed by heterotrophs, from either an endogenous or ex- ogenous source, may be dependent on the metabolic state of the organisms and the specific environmental conditions.In their early work, Romanenko et al. (1972) suggested that the rate of heterotrophic anaplerotic fixation of DIC is strictly proportional to the heterotrophic bacterial carbon production.Since then, a number of factors have been identified potentially influencing the relative contribution of anaplerotic and other non-autotrophic CO 2 fixation reactions on biomass production.In laboratory experiments with the bacterial strain Thiobacillus novellus, for example, a higher amount of CO 2 was fixed under nutrient-limited conditions (Perez and Matin, 1982).Moreover, mixotrophic bacterial strains fixed more DIC compared to those grown autotrophically (Perez and Matin, 1982).Fungi fixed relatively more CO 2 at lower organic carbon (glucose and maltose) concentrations (Schinner et al., 1982).In conclusion, the degree of heterotrophic CO 2 fixation highly depended on the availability of easily degradable organic carbon sources (Schinner et al., 1982).
Studies on the possible relationship between heterotrophic DIC fixation and the activity of prokaryotic cells revealed contradicting results.While Roslev et al. (2004) mentioned actively growing cells fix more DIC than resting cells, Merlin et al. (2003) report enhanced uptake of DIC by heterotrophic bacteria during slow growth and starvation.A relationship between DIC and heterotrophic bacterial production has been reported frequently as exemplified below.

Respiration and non-autotrophic CO 2 fixation
The production of CO 2 via respiration and the parallel fixation of CO 2 in heterotrophs take place simultaneously.The heterotrophic fixation of CO 2 is thus generally considered a back-reaction; i.e., part of the originally produced CO 2 from respiration is re-assimilated.Following this line of arguments, the more reduced an organic substrate is the less CO 2 is released (Fig. 2).The heterotrophic fixation of DIC does not necessarily lead to a net carbon biomass production; however, if microbes oxidize geogenic methane, this would result in a net carbon biomass production.Experimentally it is difficult to differentiate respiratory CO 2 flux from concurrent anaplerotic CO 2 fixation.As a consequence, there https://doi.org/10.5194/bg-18-3689-2021 Biogeosciences, 18, 3689-3700, 2021 are numerous experiments and field studies determining dark CO 2 fixation, but only a few studies quantified the assimilation of DIC by non-autotrophs.
Respiration in aquatic systems is frequently determined via the consumption of dissolved oxygen (Robinson and Williams, 2005), potentially underestimating the carbon use efficiency of heterotrophs.Depending on the substrate, the respiration quotient ( CO 2 / − O 2 ) varies between 0.7-1.3(Robinson, 2019), leading to an error between 20 % and 40 % with regard to CO 2 production from respiration.Moreover, the respiration quotient also varies because other oxygenconsuming processes are potentially taking place simultaneously (e.g.nitrification) (Robinson, 2019).For instance, it is 138 O 2 for 106 CO 2 for ideal Redfield-type organic matter and 150 O 2 for 106 CO 2 for more realistic marine organic matter (Fraga et al., 1998;Paulmier et al., 2009).Calculations based on a study on temperate forest soils revealed a reduction of overall CO 2 emissions due to dark CO 2 fixation by mainly heterotrophic microbes (Akinyede et al., 2020).Collectively, with respect to C cycling, heterotrophic CO 2 fixation and the carbon flux from the inorganic pool into heterotrophic biomass can be regarded as a process more important than hitherto assumed.

CO 2 fixation in habitats dominated by heterotrophs
In contrast to sunlit habitats, where photoautotrophs make up a significant portion of the total biomass and photosynthesis is of major importance in carbon cycling, heterotrophs and chemolithoautotrophs represent the prevailing biota in the "dark habitats", i.e., soils, subsurface environments, and the deep sea.These dark environments, although characterized by disproportionally lower biological activity, exceed their photic counterparts in both volume and biomass.In the oceans, the deep sea (below 200 m) exceeds the sunlit surface layer by a factor of 18 in volume and, remarkably, by a factor of 2 in biomass (Arístegui et al., 2009).Therefore, the so-called "dark CO 2 fixation" not only occurs in specific "hot spots" on the seafloor (hydrothermal vents, cold seeps, and mud volcanoes), or in anoxic waters, but also throughout the entire oxygenated "dark" water column (Reinthaler et al., 2010;Yakimov et al., 2014).In limnic environments, the dark groundwater ecosystems outnumber surface waters 100-fold in terms of water volume (Danielopol et al., 2003), and similarly, soils are also, with the exception of their surface, exclusively dark habitats.
Yet, heterotrophic CO 2 fixation does not occur only in the dark environments since heterotrophs are also found in the photic zone.This is particularly relevant in the ocean because the photic zone is where the highest biomass concentrations are found.Recently, it has been estimated that the inclusion of dark CO 2 fixation (integrated over the euphotic layer, 0-150 m depth) would increase oceanic primary production estimates by 2.5 %-22 % (Baltar et al., 2019).A similar situation might be assumed for surface inland waters; however, global estimations are missing so far.
As indicated, part of the dark CO 2 fixation in oceans has been attributed to chemolithoautotrophic archaea (Wuchter et al., 2003;Ingalls et al., 2006), obtaining the energy required for the endergonic carboxylation through the oxidation of reduced inorganic compounds, such as ammonia or hydrogen sulfide (Swan et al., 2011;Zhang et al., 2020).A total annual chemolithoautotrophic CO 2 fixation rate of 0.77 Pg C was calculated for the oceans (Middelburg, 2011).The observed fluxes of the reduced inorganic compounds available as energy sources, however, seem largely insufficient to explain the relatively high dark CO 2 fixation rates (Overbeck, 1979;Tuttle and Jannasch, 1979;Baltar et al., 2010;Reinthaler et al., 2010;Herndl and Reinthaler, 2013).In some cases, the supply rates of the reduced inorganic compounds used as an energy source explain less than 40 % of the observed dark CO 2 fixation rates (Zopfi et al., 2001).Recently, chemoautotrophic nitrification was estimated to explain < 13 % of the dark CO 2 fixation (integrated over the euphotic zone), with the rest coming from either heterotrophic DIC fixation or other chemoautotrophic processes (Baltar and Herndl, 2019).
The potential energy sources for the unexplained proportion of the dark CO 2 fixation remain enigmatic.Possible explanations could be either an underestimation of the supply rates of reduced inorganic compounds or the uptake of CO 2 by heterotrophic organisms (Zopfi et al., 2001;Baltar et al., 2019).In the surface ocean in particular, DIC incorporation via anaplerotic reactions might play an important role in compensating for metabolic imbalances in marine bacteria under oligotrophic conditions, contributing > 30 % of the carbon incorporated into biomass (González et al., 2008;Palovaara et al., 2014).Evidence for the latter comes from experiments with Arctic seawater, which exhibited high DIC fixation rates (0.5-2.5 µg C L −1 d −1 ) correlating with heterotrophic bacterial production (Alonso-Sáez et al., 2010).Using different molecular tools, DIC uptake was attributed mainly to heterotrophic Gammaand Betaproteobacteria rather than to typical chemoautotrophs, thus showing that chemolithoautotrophs were not the main drivers of CO 2 fixation in this habitat (Alonso-Sáez et al., 2010).Further evidence comes from the genome of Polaribacter sp.MED152, a representative of Bacteroidetes, which typically comprise about 10 %-20 % of the prokaryotic abundance in seawater (González et al., 2008).A unique combination of membrane transporters and carboxylases in these organisms indicates the importance of anaplerosis in addition to other DIC fixation pathways (González et al., 2008).If the heterotrophic metabolism of bacteria is suddenly intensified (e.g., after an input of organic matter), dark DIC fixation rates and the expression of transcripts associated with key anaplerotic enzymes increase proportionally (Baltar et al., 2016).As men-tioned above, contradicting results were obtained on the relationship between heterotrophic CO 2 fixation and the availability of organic matter.A few studies suggest a relative increase in dark DIC fixation in oligotrophic habitats harboring slow-growing or starving bacterial populations (Perez and Matin, 1982;Schinner et al., 1982;Merlin et al., 2003;Alonso-Sáez et al., 2010;Santoro et al., 2013).Considering the slow community-wide specific growth rates of heterotrophic bacteria in oligotrophic and/or cold waters, such as the marine aphotic zone, the Arctic Ocean, deep sea sediments, groundwater systems and the terrestrial subsurface, alpine limnic systems, and deep-lake sediments, enhanced anaplerotic DIC uptake can be expected.However, there is also evidence for the stimulation of dark DIC fixation in response to organic matter enrichment in different types of soils (Miltner et al., 2005;Šantrůčková et al., 2018).Hence, these contradictory findings require further, more systematic research.
Other environmental factors that may influence dark DIC fixation include the concentrations of CO 2 and bicarbonate as inorganic carbon sources.An increase in the CO 2 concentration may shift the equilibrium of the carboxylationdecarboxylation reactions increasing CO 2 fixation.Elevated partial pressure of CO 2 might stimulate dark DIC fixation.In temperate forest soils, rates of dark microbial CO 2 fixation were positive correlated with the CO 2 concentration (Spohn et al., 2019).Similarly, with increasing CO 2 concentrations, higher dark DIC fixation was observed in wetland soils affected by subcrustal CO 2 degassing (Beulig et al., 2015).Here, besides known chemoautotrophs, CO 2 fixation via anaplerotic reactions was shown for putatively heterotrophs, i.e., subdivision 1 Acidobacteriaceae, lacking enzymatic pathways for autotrophic CO 2 fixation (Beulig et al., 2015).In experiments with two marine heterotrophic bacterial isolates, elevation of CO 2 concentration provoked an increase in CO 2 fixation along with a decrease in respiration (Teiro et al., 2012).Thus, we may assume that a rise in CO 2 concentrations and CO 2 -induced geochemical changes will alter carbon turnover in affected ecosystems, with dark DIC fixation and anaplerotic reactions becoming more important.
5 Quantitative estimates of heterotrophic CO 2 fixation in different environments

Quantification of heterotrophic DIC fixation
It is difficult to properly quantify heterotrophic fixation of inorganic carbon in the environment.Not surprisingly, quantitative data almost exclusively originate from laboratory experiments using cultures and tissues in combination with carbon isotopic labeling (e.g.Spona-Friedl et al., 2020).Field studies generally report on dark carbon fixation, including the activity of chemoautotrophs and heterotrophs.So far, evidence for the significant contribution of heterotrophic CO 2 https://doi.org/10.5194/bg-18-3689-2021Biogeosciences, 18, 3689-3700, 2021 Table 1.Dissolved inorganic carbon (DIC) assimilation rates from a range of aquatic (marine and limnic) and soil environments.Dark carbon fixation (DCF) is shown as fraction of either bacterial heterotrophic production (BP) or respiration.Original data were converted to similar units whenever possible to allow comparison.Aquatic  Anaplerotic enzymes comprised the majority of carboxylase genes fixation, as highlighted for selected studies in Table 1, is based on additional measures complementing the quantification of dark carbon fixation, i.e. molecular microbial community analysis occasionally including the quantitative assessment of functional genes involved in carbon fixation and characterization of environmental conditions.In fact, significant dark carbon fixation (i) in the obvious absence of chemoautotrophs and related genes or (ii) in well-oxygenated environments that lack potential electron donors (e.g.H 2 , H 2 S, NH 4 ) led authors to conclude on the quantitative importance of heterotrophic carbon fixation (e.g.Miltner et al., 2005;Alonso-Saéz et al., 2010;Yakimov et al., 2014;Šantrůčková et al., 2018;Akinyede et al., 2020).Individual studies succeeded to follow the label of inorganic carbon in environmental samples into amino acids of microorganisms (Spohn et al., 2019;Spona-Friedl et al., 2020).In the future, the combined application of metabolomics and isotope tracing may help further develop this field of research.

Heterotrophic CO 2 fixation in different habitats
Measurements of dark DIC fixation with strong evidence of a significant contribution of heterotrophic assimilation of DIC are scarce.In Table 1, we provide a compilation of studies conducted in soils and marine and limnic ecosystems.Where possible, we compared dark DIC fixation rates with heterotrophic activity.In marine and limnic systems, heterotrophic carbon production as a widely applied activity measurement was used.In soils, we compared dark DIC fixation rates with respiration, i.e., CO 2 production.Dark DIC fixation rates in different marine systems range between 0.1 and 206 µg C L −1 d −1 , with the highest values found in a eutrophic lagoon and the lowest values in the deep waters of the Mediterranean Sea (Table 1).Data from limnic systems originate from lake sediments with dark DIC fixation rates between 0.12 and 48 mg C m −2 d −1 (Table 1).Projecting these numbers to only the top 10 cm of sediment in the different lakes studied (which is a gross simplification), values of 1.2-480 µg C L −1 sediment per day are obtained.When compared to rates of bacterial carbon production, dark DIC fixation rates in these habitats accounted for a considerable fraction of total carbon assimilation, occasionally even exceeding it (Table 1).In soils, the dark DIC fixation rates, which were attributed mainly to the activity of heterotrophs, amounted to 0.04 %-39 % of the overall respiration rate (Table 1).Dark DIC fixation rates range from 36 ng C to 23.6 µg C g −1 d −1 covering 3 orders of magnitude (Table 1).The contribution of heterotrophically fixed DIC to biomass carbon of microbes ranged from 0.2 %-1.1 % in temperate forest soil (Akinyede et al., 2020), 0.2 %-4.6 % in temperate forest and field soils (Šantrůčková et al., 2005), to 7 % in arable soil (Miltner et al., 2004).Šantrůčková et al. (2005) estimated the overall heterotrophic CO 2 fixation to be even higher, i.e., 1.9 %-11.3 % taking into account that the labile fraction of the biodegradable organic carbon re-sulted from metabolites released by spilling reactions of microorganisms due to a limitation in inorganic nutrients or due to the presence of highly reduced energy-rich carbon sources (e.g.Tempest and Neijssel, 1992).A contribution of heterotrophic CO 2 fixation to biomass carbon of 6.5 ± 2.8 % was found in drinking water biofilms and activated sludge (Roslev et al., 2004).

Carbon biomass stock originating from heterotrophic CO 2 fixation
While it is difficult to derive global estimations from the few studies that measured heterotrophic CO 2 fixation rates in marine, limnic, and terrestrial ecosystems, we may use a conservative approach assuming that at least 1 %-5 % of carbon biomass of all heterotrophs originates from anaplerotic DIC fixation.Earth's total living biomass is estimated to amount to about 499-738 Pg C, of which approx.451-653 Pg C is photoautotrophic biomass (Bar-On et al., 2018).Heterotrophic biomass thus contributes 47-85 Pg C (Table S1 in the Supplement).The uncertainties of the estimates of heterotrophic biomass of the terrestrial subsurface, however, are high (Whitman et al., 1998;McMahon and Parnell, 2014;Bar-On et al., 2018).Nevertheless, following this line of evidence, anaplerotic CO 2 fixation contributes between 0.5-5 Pg C to the living biomass.

Carbon flux related to heterotrophic CO 2 fixation
In terms of annual global heterotrophic production rates, oceans and the terrestrial subsurface (including soils) are the main habitats of heterotrophic CO 2 fixation (Cole et al., 1988;Magnabosco et al., 2018)  Our estimates, as said, are subject to a high uncertainty, which, on the one hand, results from the dependency of the extent of heterotrophic CO 2 fixation on the organic carbon oxidized and, on the other hand, on the predominant environmental conditions.Moreover, data on terrestrial and marine subsurface environments, although large in dimension, are scarce.For these environments, no detailed information on the abundance, growth (yield), and metabolic activity of microbial communities is available, particularly with increasing depth.Most of the deeper subsurface environments, even when harboring considerable living biomass, do not https://doi.org/10.5194/bg-18-3689-2021 Biogeosciences, 18, 3689-3700, 2021 participate in the global carbon cycle on short and medium timescales (years to decades), but rather on centennial to geological timescales.Nevertheless, in order to provide a first estimate and to be able to roughly evaluate the relevance of heterotrophic CO 2 fixation for all habitats, we adopted this conservative approach (see also Tables S1 and S2).

Conclusions
Current models of carbon cycling and carbon sequestration (Gruber et al., 2004;Le Quéré et al., 2009) do not account for heterotrophic CO 2 fixation.Despite the uncertainties in the data on heterotrophic biomass and production rates for some habitats (e.g. the terrestrial subsurface), the numbers presented here represent the first attempt to quantify the global contribution and relevance of heterotrophic CO 2 fixation to carbon cycling.Our results indicate that heterotrophs significantly contribute to global CO 2 fixation -especially (although not restricted to) in habitats experiencing elevated CO 2 concentrations and/or lacking a sufficient supply of degradable organic carbon.In specific environments, this may explain the mismatch between autotrophic C input, consumption, and sequestration as has been observed in marine systems (Baltar and Herndl, 2019;Burd et al., 2010;Reinthaler et al., 2010;Morán et al., 2007;Hoppe et al., 2002;Tait and Schiel, 2013).Particularly in aphotic habitats (which outnumber the photic habitats in both size and volume) such as the dark ocean, subseafloor sediments, soils, and the sediments and rocks of the terrestrial subsurface (Miltner et al., 2004(Miltner et al., , 2005;;Yakimov et al., 2014;Wegener et al., 2012), carbon cycling needs to be re-evaluated taking into account anaplerotic CO 2 fixation and other inorganic carbon uptake pathways in heterotrophs.In subseafloor sediments, wetlands, and marshes, as well as in other habitats where methane oxidation is a key process, a large fraction (10 %-50 %) of heterotrophic biomass potentially originates from heterotrophic DIC fixation.Recently, a timeseries study showed a tendency towards higher ratios of dark to light DIC fixation in the top half of the euphotic layer (0-65 m) in the years 2012-2019 than in the preceding years (data started in 1989), which was linked to oceanographic changes (i.e., a deepening of the mixed zone) (Baltar et al., 2019).Moreover, the metabolic theory of ecology posits that heterotrophic metabolism increases more than gross primary production in the ocean in response to warming (see Baltar et al., 2019, and reference therein), which might also make heterotrophic DIC fixation relatively more important in a warmer ocean.In light of global warming leading to an extensive thawing of permafrost soils and providing new habitats for methanotrophs, these processes are expected to become more important in the future.Hence, the potential contribution of heterotrophic CO 2 fixation under climate change conditions clearly deserves further investigation.
Data availability.All data that have been used for our calculations are available in Tables S1 and S2 in the Supplement.No further data sets were used in this article.
Author contributions.AB, ME, and CG conceived the idea for the manuscript.AB, GJH, and CG wrote the manuscript.MSF, ME, MA, FB, and TR substantially commented on and edited the manuscript.MA, MSF, and CG did the literature search on available global carbon data.CG and MA performed the estimation of heterotrophic CO 2 fixation on a global scale.
Competing interests.The authors declare that they have no conflict of interest.

Figure 2 .
Figure 2. Anaplerotic CO 2 fixation contributes 1 %-8 % of carbon biomass (indicated by the grey band) in heterotrophic cells.Dependent to the organism and in relation to the uptake of the individual organic compounds and their entry into the TCA cycle and central metabolic pathways the relative amount of inorganic carbon assimilated varies, as highlighted by the red arrows.See examples for malic and lactic acid.With organic carbon sources more reduced than the organism's biomass (dashed line) further carboxylation reactions are induced (indicated by black line), increasing the overall carbon contribution from CO 2 beyond anaplerosis (grey band).In methanotrophs, 50 % of the cell's carbon may originate from CO 2 fixation.For further explanations, see text.