All living organisms need energy, C sources, and nutrients to accommodate their requirements for maintenance metabolism, growth, and activity. In soils, some microorganisms are photolithoautotrophs (green algae, cyanobacteria) or chemolithoautotrophs that use inorganic molecules, e.g. NH₄ ⁺ or Fe²⁺ as energy sources and electron donors as well as CO₂ as C source; for more details see (Lengler et al., 1999; Schink, 2006). However, most of the microorganisms in soil are chemoorganoheterotrophs and organic molecules are their C and energy sources. The C of these organic substrates can be shared between catabolism (= degradation) and anabolism with biomolecule synthesis and eventually growth. Thereby the catabolism provides the energy for the anabolism (Figure 1). Under starvation or unbalanced elemental supply conditions in soil, anabolic C turnover can be decoupled from catabolic processes, resulting in variable C use from different substrates. However, no anabolism is possible without energy gain from catabolism (Russel and Cook, 1995; Lengler et al., 1999). Degradation processes in catabolism apply irrespective of the origin of the substrates for both plant materials and microbial necromass. The turnover of microbial bio- and necromass enables element cycling of N and P through microbial food webs (Kästner and Miltner, 2018). Productive degradation of a substrate, combined with cycling of biomass building blocks, can increase the overall amount of biomass produced.

Energy is the driver of all biotic processes in nature and energy metabolism depends on redox reactions that couple two proceeding half-reactions (oxidation and reduction). The total energy gain (ΔG, Gibbs energy) is the energy balance of the half-reactions in heterotrophic organisms: 1) the oxidation of C substrates, often to CO₂, and 2) the transfer of released electrons to terminal electron acceptors in respiration processes, e.g. the reduction of O₂ to H₂O (Russel and Cook, 1995; Schink, 2006). The energy gain from the oxidation of substrates varies with different electron acceptors; O₂ as electron acceptor sustains the highest turnover rates of organic matter (Schink, 2006). Microbial activity generally reduces the redox potential and O₂ content in the vicinity of the cells. Alternative electron acceptors such as NO₃ ⁻, Fe³⁺, Mn^4+/3+, SO₄ ^2- can also be used, but with decreasing energy gains due to lower redox potentials. Anoxic electron acceptors result in various degradation efficiencies of molecule classes, which results in selective stabilization (Keiluweit et al., 2017). Thermodynamics can thus provide mechanistic explanations for SOM formation and turnover and for understanding soil processes in general.

Microbial growth and biomass formation by anabolism in soils depends on many factors such as 1) soil properties, texture, and environmental conditions (pH, clay and SOM content as well as humidity, aeration, temperature), 2) the presence and general activity of potential litter degraders in microbial communities, and 3) the availability of nutrients, electron acceptors, and substrates (Gavrilescu, 2005; Waring et al., 2020). Bioavailability is controlled by the characteristics of the substrate (water solubility, molecular size, K_oc/K_ow values) as well as their interaction with other components of the soil matrix, e.g., dissolution, sorption, aggregation, and entrapment (Bollag and Liu, 1990; Sims et al., 1990; Gavrilescu, 2005).

Under optimal conditions with ideal water contents, electron acceptor availability, nutrient and substrate supply, the growth rates of microorganisms depend on available C and energy substrate concentrations. For growth, each cell needs to divert a minimum substrate flux in order to maintain metabolism, cell integrity, and survival (the maintenance threshold (van Uden, 1967; van Bodegom, 2007)). Thus, minimum substrate and energy fluxes significantly different from zero are needed, even if cells do not grow. Low concentrations of substrates or low turnover rates due to limited supply conditions may not provide sufficient energy for this maintenance metabolism, resulting in dormancy or spore formation and eventually in cell death (Kovarova and Egli, 1998). For a long time it was thought that the maintenance rates of various bacteria as well as death rates and growth yields were nearly constant over wide substrate ranges, but this view has changed (van Bodegom, 2007). More detailed examinations with a model bacterium showed that fast-growing bacteria die faster upon substrate deprivation and have lower biomass yields due to increased maintenance requirements (Biselli et al., 2020). This means that growth rates are positively correlated to maintenance needs and death rates; however, they are negatively correlated with biomass yields and increasing growth rates can thus result in decreasing biomass yields relative to substrate uptake. Higher growth rates thus can lower C retention in biomass resulting in highly flexible microbial C use in soil. Slow-growing cells on the other hand survive longer due to lower maintenance requirements (Russel and Cook, 1995) and have an evolutionary advantage in environments where substrate supply varies over time (Kovarova and Egli, 1998; van Bodegom, 2007; Biselli et al., 2020). This advantage is based on lower maintenance needs and thus the ability to survive longer (Kovarova and Egli, 1998; van Bodegom, 2007; Biselli et al., 2020). Not surprisingly, the maintenance requirements and survival of organisms also depend on the recycling yield from dead cells in microbial cultures. Organisms usually feed on their deceased neighbors and the recycling yield is exponentially related to the growth rate (Biselli et al., 2020). Under optimal growth conditions a recycling yield of 12% can be obtained. In other words, only eight dead cells are needed to produce one new cell, which supports the hypothesis that microbes are capable of generally mining microbial necromass. The requirements are much lower under more limited conditions. This mining can explain why microbes in deeper soil layers can survive by merely feeding on the necromass seeping downwards; the availability of these resources directly regulates the C use (Ludwig et al., 2015; Ni et al., 2020).

With decreasing concentrations of degradable SOM compounds, the overall growth in soil decreases (Kästner and Miltner, 2018; Zheng et al., 2019). However, the activity of microorganisms is determined by the energy content of the substrates, the availability of nutrients, terminal electron acceptors as well as soil water contents, and typically by the most limiting factor (Russel and Cook, 1995; Lengler et al., 1999; Waring et al., 2020). For optimum growth, all factors required by the soil microorganisms have to match their needs. This does not only include C but also other nutrients in the required stoichiometric ratios. The growth of soil microorganisms is often limited by one of these factors, which then controls the growth yield and energy use efficiency (EUE = energy retained in the system, here in biomass or SOM) (Harris et al., 2012). The ratio of C used for biomass production (= growth) in anabolism over total substrate C use (= C used for growth + respired C) has been defined as yield coefficients (Y_X/S) in microbiology (Schink, 2006) which is similar to C use efficiency (CUE) (del Giorgio and Cole, 1998).

The organic matter of the photoauthotrophic primary producers enters the soil as litter and root exudates (Schmidt et al., 2011; Lehmann and Kleber, 2015b; Janzen, 2015; Paul, 2016). Litter mostly consists of macroscopically visible aggregated plant-derived biopolymers (such as starch, cellulose, and lignin); for details see (Kögel-Knabner, 2017) and the references therein. Due to their large molecular size, these compounds cannot be taken up by microorganisms (with cell sizes of a just a few μm), because the transport across the cell envelope is considered to be limited to smaller molecules (<600–1,000 Da) (Ekschmitt et al., 2008; Lehmann and Kleber, 2015a). Similarly, microbial biomass and their aggregates of proteins, nucleic acids, cell envelope fragments, and extracellular polymeric substances are also macromolecules, but of a much smaller size in the order of nm. All of these macromolecules need to be depolymerized (often hydrolyzed) to oligomers outside of microbial cells by extracellular enzymes before they can be taken up as substrates (Schimel and Weintraub, 2003; Kästner and Miltner, 2018). Microorganisms producing extracellular enzymes were shown to have a lower growth yield than non-producers (Malik et al., 2019). The accumulation of microbial necromass alters the relative C and N limitation during litter decomposition and accelerates N cycling, ultimately resulting in increased SOM content. Such effects partly decouple C and N use and the spatial heterogeneity may explain different C and N turnover dynamics and stabilization in SOM. The resulting monomers and oligomers can then be taken up by microorganisms and used for cell maintenance and growth.

Due to their rapid consumption, such low-molecular-weight compounds are only present at very low steady-state concentrations in the soil, although the flux into the cells may be considerable (Schimel and Schaeffer, 2012). Sorption reactions often take place very quickly and can reduce the availability of substrates for microorganisms, resulting in a competition between uptake with degradation by microorganisms and external reactions. As bioavailability of C substrates and nutrients is often limited in soil, starving microorganisms are eager to take up whatever they are able to access and able to degrade (Schimel and Schaeffer, 2012; Waring et al., 2020). Described transport mechanisms for the uptake of organic molecules into microbes are (Lengler et al., 1999; Madigan et al., 2014): 1) diffusion, or 2) porin-mediated transporter systems, which are typically transmembrane tunnel proteins in the cell membrane that can be associated to electron transfer, acceptor molecules, or a phosphorylation process. The abundance and properties of such porines can be modulated by the microorganisms and depend on environmental and growth conditions as well as exposition to substrates and chemicals (Denyer and Maillard, 2002). In addition, facilitated transport of molecules into degrading microbial cells by unknown mechanisms related to carrier molecules, e.g. by the presence of humic acids or particularly after compost addition, has also been described (Li et al., 2021; Adam et al.). Even sorbed compounds can be degraded (Feng and Boyd, 2008) and for Gram-positive bacteria the uptake of N-containing compounds of much higher molecular size, e.g. proteins even from sites of strong sorption, has been observed (Enggrob et al., 2020).

Temperature, water and O₂ contents of a soil are all important regulators of plant primary production, soil microbial activity, and energy use, controlling the input and compound turnover in the soil system. Access to water is vital for bacterial and fungal biodegradation and growth (Moyano et al., 2013; Rath et al., 2017) and a lack of bio-accessible water leads to limited depolymerization and degradation in general (Masoom et al., 2016). In addition, O₂ limitations may be the least understood regulators of SOM turnover (Keiluweit et al., 2017). This is also valid for microbial growth and necromass stabilization related to redox cycling. The importance of temporary anaerobic conditions for SOM accumulation is underlined by results from wet rice-cropping in paddy soils, which show a profound enrichment of SOM over 2000 years (Kalbitz et al., 2013).

Moreover, self-organization and interaction dynamics in decomposer communities are generally assumed in soils and can result in increased biomass amounts as SOM is thought to be utilized by functionally redundant soil-specific microbial communities (Rillig et al., 2015; Fierer, 2017). Based on the modelling approach of Allison (2005), Kaiser et al. (2014); Kaiser et al. (2015) showed that SOM development depends not only on the stoichiometric response of microbes to a high C:N ratio substrate but also on the community dynamics. They showed that the activity of microbes producing extracellular enzymes (“decomposers”) and microbes exploiting the catalytic activities of others (“cheaters”) lead to the regulation of SOM turnover. The presence of “cheaters” increased N retention by down-regulation of the ratio of extracellular enzymes to total microbial biomass, resulting in increased bio- and necromass formation. This resulted in increased N-rich necromass accumulation, increased C retention, decreased turnover rates, and an increase in spatial heterogeneity of microorganisms and necromass distribution.

In spite of the importance of energy fluxes through the soil systems, thermodynamics have only partly been considered for soil systems (Bosatta and Agren, 1999; Lueders et al., 2004; Harris et al., 2012; Di Lonardo et al., 2017; Manzoni et al., 2021), but have been applied predominantly to technological, chemical, and biochemical processes, such as the production of proteins, and in order to determine the energy use, heat production, and nutrient requirements (Westerhoff et al., 1982; McCarty, 2007; Xiao and VanBriesen, 2008; Kleerebezem and Van Loosdrecht, 2010). The determination of thermodynamic state variables has been applied to soil or natural organic matter processes only in a handful of cases (Barros and Feijóo, 2003; Barros et al., 2007; Herrmann et al., 2014; Barros et al., 2016; Boye et al., 2018). Currie (2003) identified the link between energy and C content in the soil by combining the principle of energy balance with a biogeochemical process model and using calorimetric analysis of different litter materials and soil samples for parameterization of the turnover processes. However, the main obstacle for considering energy fluxes was the difficulty in finding suitable experimental approaches that link energy to matter fluxes in soil systems. For soil they have to be related to complex and diverse energy-consuming processes and including the formation of microbial biomass, necromass, and subsequently SOM (Chakrawal et al., 2020; Barros, 2021).

Thermodynamic properties of low-molecular-weight organic compounds have been related to their elemental composition and the nominal oxidation state of C (NOSC) in organic compounds (including SOM) and were suggested as a proxy for the potential release of Gibbs energy during the oxidative degradation of these compounds (LaRowe and van Cappellen, 2011). The NOSC was then used to explain qualitatively the linkage between organic C transformation, soil microbial communities and land use/management (Keiluweit et al., 2017). However, NOSC was not found to explain the preferential substrate use of various organic compounds in pure culture experiments (Cyle et al., 2020). Emphasizing the thermodynamic potential factors as a driving force for microbial metabolism (Ji and Bethke, 2007), thermodynamic constraints were shown to lead to a limited degradation of highly reduced compounds (e.g., lipids) under anoxic conditions (Keiluweit et al., 2017), resulting in the relative enrichment of such compounds in anoxic microsites of soil aggregates. Thermodynamic aspects thus play a much more important role in SOM protection and stabilization than was previously assumed.

Quantitative approaches using thermodynamics in the soil are currently limited to defined organic compounds. The potential growth yield of microbes feeding on a certain compound has been modelled in order to predict their turnover to biomass in soils (Brock et al., 2017; Trapp et al., 2018), including stabilization in SOM. The Microbial Turnover to Biomass (MTB) model is based on the estimation of the energy available for biomass formation (growth) obtained from the catabolic reactions of a substrate. The model calculates the electron transfers that can be exploited by microorganisms from the substrate to the respective electron acceptors for a specific turnover equation (Brock et al., 2017; Trapp et al., 2018); it predicts the resulting potential microbial yields and enables the formation of biomass and necromass to be related to the CO₂ released, hence it enables a direct link to potential C use in soil. The approach can be applied for many organic compounds in order to predict their potential contribution to microbial biomass formation and ultimately the stabilization of organic compounds in the soil.

Considering the thermodynamic issues, we can conclude that energy is needed for SOM formation through microbial growth and the decay cycle in which some energy is retained within the necromass. A comparison of the total Gibbs energy released from an oxidation reaction with the potential energy available for microbial growth provides an estimate of the actual energy retention within microbial biomass and later necromass in a soil in relation to the release as heat and may provide a sound basis for further research approaches for improving SOM contents.

There are still huge gaps in our understanding of C conversion and microbial CUE in soils, which is defined as the C stored in soil per substrate consumed (del Giorgio and Cole, 1998; Mooshammer et al., 2014; Nunan et al., 2015; Spohn, 2016; Spohn et al., 2016; Manzoni et al., 2018; Geyer et al., 2019; Manzoni et al., 2021). CUE will always be lower than 1 and the empirically determined upper limit is 0.6 (Roels, 1980); it is not a constant value and depends on the microbial physiology and the environmental conditions (temperature, water content, electron acceptor availability). CUE was often expected to be constant; however, the related CUE in complex environmental systems must be calculated based upon the sum of all consumed substrates and thus it may be variable (Manzoni et al., 2021). CUE is a key parameter that is closely linked with the biotic energy use principles on all trophic levels and that varies with the input of C and energy of the substrates and the specific environmental conditions. Reports consider various influencing factors such as depolymerization of plant matter, substrate elemental composition (stoichiometry), substrate molecular structure, and nutrient availability (Liu et al., 2016; Spohn et al., 2016; Takriti et al., 2018). However, microbial necromass as a substrate and nutrient resource also plays a significant role. Due to the variation and environmental conditions of soils, the CUE was observed in the range of 0.3–0.55 (Sinsabaugh et al., 2016; Soares and Rousk, 2019). Substantial differences in CUE were determined by various methods targeting different aspects of microbial metabolism, e.g. general growth, biomarker formation, energy gain by ATP formation, etc. (Geyer et al., 2019). CUE based on the assimilation of radiolabeled thymine and acetate often provide relatively low values (0.03–0.3) (Soares and Rousk, 2019). These differences complicate comparisons of CUE reported in different studies.

CUE in soils shows an expected high sensitivity and negative correlation to temperature (Manzoni et al., 2012). The response of CUE to short-term changes in temperature and moisture as well as O₂ content was found to depend mainly on microbial growth response and not on respiration responses in various soils and land use types (Zheng et al., 2019). The C:N ratios of plant inputs are negatively correlated to CUE; the highest CUE in terrestrial ecosystems are found with plant input material C:N ratios around 10, which are in the range of microbial biomass C/N (Manzoni et al., 2012). Such materials meet the stoichiometric requirements for microbial biomass production and thus favor microbial growth. CUE declines as resource C:N moves further away from biomass stoichiometry but also depends on the availability of inorganic and organic N (Manzoni et al., 2008; Manzoni et al., 2021). Under N-limited conditions, CUE is reduced and nitrogen use efficiency (NUE) is maximized, whereas under non-limited conditions NUE decreases (Mooshammer et al., 2014). Conversely, C:N:P stoichiometry regulates SOM mineralization; N and P provision increases biomass formation and lowers CO₂ emission (Wei et al., 2020). In addition, soil warming increases growth, decreases CUE and increases microbial turnover to CO₂, finally affecting SOM dynamics, see Figure 2 (Li et al., 2019b). With the focus on necromass, high CUE in soils occurred by necromass retention in relation to the substrate consumption for biomass production, and is thus a matter of microbial physiology as well as of matter and energy fluxes. Therefore, we can hypothesize that rapid degradation processes of a pulsed substrate input and the related microbial growth result in peak energy and C fluxes through the soil, whereas slow processes result in much broader flux behavior with higher overall C and energy retention (Figure 2). This results in higher cumulative microbial biomass formation as the organisms are able to meet their maintenance requirements over longer periods of time, translating into more necromass and SOM formation.

In addition to C and energy, microorganisms need other elements, such as H, O, N, etc. at appropriate stoichiometric ratios for homoeostatic growth. The elemental composition of microbial biomass may vary from C₄H₇O_1,5 N (Lengler et al., 1999) to C₅H₇O₂N (Christensen and McCarty, 1975) and C₅H₈O_0,8 N (Rittmann and McCarty, 2001), with C:N ratios of about 4-5 in microbial cultures. In addition, small amounts of P and S as well as trace elements (e.g. Fe, K, Mg, Ca) must be available (Lengler et al., 1999). Soil microbial biomass C:N ratios were reported to be higher, ranging from 4.5 to 12.5, with 7 as a generally accepted average (Xu et al., 2013; Mooshammer et al., 2014). Soil microorganisms usually grow on plant litter material, which has higher C:N ratios than microbial biomass. The microorganisms have several options to meet their stoichiometric requirements, e.g. they could preferentially degrade nutrient-rich compounds from plant matter, mine SOM for specific elements, or utilize other substrates with the excess C released into the environment. In addition, living microorganisms can also recycle cell internal materials (Spohn et al., 2016; Capek et al., 2021). The elemental needs are the key element of the theory of ecological stoichiometry and are one of the potential mechanisms for a variation in CUE and priming effects in soils (Nunan et al., 2015; Spohn et al., 2016). From turnover analyses of organic compounds in the soil we know that multiple C sources are often used by microbes in order to obtain the optimum stoichiometry by “mixing” substrate molecules (Kovarova and Egli, 1998; Brock et al., 2019; Cyle et al., 2020).

CUE in soil cannot be related to the transformation of a single substrate, because soil microbial communities use different C sources at the same time, which is also related to the stoichiometry needs. CUE is thus notoriously difficult to analyze related to both method and system properties (Geyer et al., 2016; Geyer et al., 2019; Li et al., 2019b). This, together with the pronounced effect of environmental boundary conditions, results in the observed variability of CUE, reflecting the plasticity of microbes reacting to their overall growth conditions (Anthony et al., 2020). The CUE concept relates C turnover to microbial growth and provides much deeper insight into the relevant processes of C dynamics when considering biomass and the resulting necromass formation as substrate and nutrient resources in soil. However, over the last decade the role of energy fluxes has hardly been given any consideration, with the main focus having been on the turnover of C, N, and P. Combined mass turnover and energy balances enable evaluation of the amount of energy derived from C sources degradation as well as the amount of energy retained in soil by microbial biomass (Barros and Feijóo, 2003; Barros, 2021).

With every microbial compound oxidation, a part of the energy is dissipated as heat, whereas the remainder may be used for driving microbial anabolism and growth. The additional use of necromass as a source of building blocks for microbial biomass (here defined as amino acids and peptides, amino sugars and peptidoglycan oligomers, fatty acids and lipids, nucleotides and nucleic acids of microbial origin) may improve growth yield, CUE and EUE considerably. For microorganisms, it is economic not to degrade all substrate molecules to central metabolites of the tricarboxylic acid cycle and then synthesize the complex biomolecules from these metabolites (see Figure 3 and also Figure 1). Recycling of building blocks from microbial necromass thus saves considerable energy, which would otherwise be needed to re-synthesize such compounds.

Based on these energy considerations, we suggest an extended mining concept of SOM decomposition, which does not only focus on the stoichiometry of elements (Mooshammer et al., 2014; Zechmeister-Boltenstern et al., 2015) but on mining of biomass building blocks that can be mobilized by living microbes. The use of biomolecules from microbial necromass improves the energy budgets and fluxes of the cells, as they do not need to consume energy for their synthesis. Microbial necromass is a good mining substrate, as the elemental and chemical composition is similar to microbial biomass and the enzymes for the depolymerization and re-polymerization are generally present in the cells. Using building blocks enables better microbial growth and biomass yields. In comparison, plant detritus has a much higher C:N ratio, is more difficult to degrade and requires depolymerization by exoenzymes, whereas plant exudates in the vicinity of roots may also directly provide building blocks.

Accordingly, a direct incorporation of labelled fatty acids from PLFA into Actinobacteria has been observed (Apostel et al., 2018). The observed higher priming effects after the addition of a single amino acid to soils compared to glucose and inorganic N (Mason-Jones et al., 2018) also support the “microbial resource mining” hypothesis, because amino acids and peptides provided as building blocks stimulate growth and additional mining. Position-specific ¹³C label showed a much higher mean residence time of ribose compared to glucose in soils, indicating the reuse or preservation of intact ribose-derived cell components in SOM (Bore et al., 2019). The addition of biomass building blocks (yeast extract) compared to glucose alone actually increased the amount of active enzymes and microbial yield in a forest soil, but did not increase the averaged growth rates or mineralization of the soil (Loeppmann et al., 2020). In addition, fungi were shown to mobilize resources from Gram-negative necromass and distribute it within the microbial food web (Zheng et al., 2021). In a rhizosphere soil an active Saccharibacterium was identified, which lacks genes for nucleotide synthesis (Starr et al., 2018) and thus depends on the uptake of complete nucleotides for DNA and RNA synthesis. Growing evidence shows that N mining can be considered to be directly targeting N-containing building blocks from microbial necromass, since a large part of N in SOM is derived from microbial products (Mooshammer et al., 2014).

However, the stabilization of microbial necromass in soil systems limits the availability for mining and is thus a competing process. Depending on the availability of nutrients, other substrates, and the biochemical effort for mining building blocks, soil microbes optimize their resource utilization in order to achieve maximum growth but with minimum energy loss (Westerhoff et al., 1982; Kovarova and Egli, 1998; Roller and Schmidt, 2015). This allows optimum growth and maintenance (= trade-off between rate and efficiency) under the given environmental conditions, and suggests that the Maximum Power Principle (Odum and Pinkerton, 1955) is relevant for the formation of SOM through microbial necromass.

Reports are inconsistent about the role of influencing factors, but the existence of optimized microbial regulation of CUE has been demonstrated along resource stoichiometry gradients (Craine et al., 2007; Manzoni et al., 2017; Manzoni et al., 2018). Increased availability of nutrients (either in organic or inorganic form) allows microorganisms to use more C for microbial growth, whereas nutrient deficiency results in less C allocation to growth and more to mineralization. The supplementation of soils with easily degradable organic compounds leads to priming effects resulting in an immediate response of the SOM turnover (Kuzyakov, 2010). Although the processes underlying priming effects are still elusive, they are presumably linked to microbial resource mining.

Predation by other organisms (Richter et al., 2019) and in particular phage lysis play important roles in determining microbial survival and death in the soil and in releasing microbial cell material (Kuzyakov and Mason-Jones, 2018). Viruses are extremely abundant in the soil (up to 10¹⁰ viral particles g⁻¹ soil (Williamson et al., 2017); and thus infection and lysis of the cells are major factors for microbial life and death in soils. After death, microbial cells start to decay and the integrity of the cell membrane is no longer given (Lengler et al., 1999). Microbial necromass can be considered as a continuum from active biomass to necromass in all stages of decay or stabilization (Kästner and Miltner, 2018). Necromass, including cell envelope fragments, is material for degradation reactions in soil, since it is intrinsically biodegradable and will be metabolized according to the same rules and principles as described for organic substrates and plant materials in general. However, due to the much smaller size of microbial biomass materials and cell envelope fragments, the depolymerization and degradation may be facilitated in comparison to plant detritus material. In addition, mining of bioavailable necromass supports the formation of new biomass, which partly ends up as necromass again within the microbial food web of necromass turnover (Lueders et al., 2006), especially in deeper soil layers (Ludwig et al., 2015; Angst et al., 2018; Ni et al., 2020). The necromass can be decomposed by enzymes in the soil matrix that are either present as exoenzymes or released by cell autolysis (Miltner et al., 2012; Schimel and Schaeffer, 2012). In addition, mechanical disruption of the residual cell envelopes by physical processes, e.g., shrinking and swelling of SOM as well as soil mixing due to bioturbation by macrofauna may also determine the fate (Bohlen et al., 2004; Basile-Doelsch et al., 2020). The ultimate fate of the necromass then depends on many factors: microbial decay, clay content, Fe or Al oxides, water content and pore system of the soil aggregates (Schimel et al., 2007).

Microbial necromass (including cell envelope fragments) can be stabilized within the soil matrix through its small size, spatial conformation, and by interaction with itself, e.g., by the formation of stacks of cell-wall fragments that may hamper degradation due to limited accessibility of the interior regions (Miltner et al., 2012; Buckeridge et al., 2020). Necromass may easily be occluded in soil aggregates and therefore not be bioavailable for degradation. Interactions of biomolecules with other organic and mineral matter can occur by ionic or hydrogen bonds or hydrophobic interactions (Galicia-Andres et al., 2021). However, various types of biomolecules are partly stabilized in SOM to different extents: proteins > bulk biomass C > lipids (e.g. PLFA) (Kindler et al., 2006; Kindler et al., 2009; Miltner et al., 2009; Spence et al., 2011; Miltner et al., 2012). An overarching stabilization process of SOM and necromass can also be caused by the formation of hydrophobic domains with a decreased chemical activity of water (Masoom et al., 2016) by drying processes, or by loss of pore water connectivity. The authors impressively provided evidence that around 75% of the necromass-derived SOM material is actually not in contact with water.

In addition, post mortem modifications of biomass components can be hypothesized, in particular for proteins, peptidoglycans and chitins, which can alter these materials and cause their stabilization, and may mask them for analysis. This could well be one reason why the high contribution of stabilized biomolecules in SOM has been overlooked for decades (Simpson et al., 2007). Necromass–necromass interactions through ionic or hydrogen bonds or hydrophobic interactions have recently been demonstrated (Buckeridge et al., 2020; Galicia-Andres et al., 2021). However, if we consider post-mortem modifications for both plant- and microbe-derived molecules, there must be some molecules resulting from these modifications. These modifications may be derived from bridging by metal cations or hydrogen bonds (Galicia-Andres et al., 2021), multiple cross-linking of proteins, peptidoglycan, lipopolysaccharides and nucleic acids, which can no longer be directly identified as the original molecule classes. For example, amorphous Fe precipitation may coat ribosomes of cellular origin. Such processes may be of high interest and particularly valid for proteins, since some authors found up to 100% of N in SOM may be present in amide bonds (Knicker, 2011).

The conversion of macroscopic macromolecular plant matter to microbial biomass causes a flux of diverse plant polymers towards small-size microbial polymers of lower diversity. Macromolecular plant matter often comprises homopolymers (e.g., starch, cellulose) as well as heteropolymers (e.g., lignin, lignocelluloses, proteins) with very high molecular weights and a size range of µm to cm (Bresinski et al., 2008). Through microbial degradation, this material is converted to microbial heteropolymers (proteins, peptidoglycan and chitines) with comparatively low molecular weights and a much smaller size (10–100 fold). The conversion of plant polymers to microbial polymers results in the “entombing effect” (Liang et al., 2017) by increasing small-scale heterogeneity, and is presumably also accompanied by an increase in entropy (von Stockar, 2010). Soil C stabilization by molecular complexity was recently conceptualized as a driving factor for the persistent behavior of generally easily degradable organic molecules (Lehmann et al., 2020), which causes the spatial separation and temporal variability of SOM. The general conversion from plant litter towards microbial OM increases small-scale molecular complexity and finally decreases the likelihood of the co-location of a substrate molecule next to the decomposer organisms, and thus the turnover rates, resulting in higher turnover times.

Experiments on the turnover of common farmyard manure in artificial soils with various mineral composition revealed relatively similar CO₂ emissions in all samples (Pronk et al., 2013). However, the highly different compositions of the developed microbial communities (Babin et al., 2013) indicated that microbial activity and utilization of complex substrates are adapted to reach the optimal CUE and are not primarily related to microbial community compositions. Although minerals and substrates strongly affect the composition of microbial communities, they obviously do not have a pronounced effect on microbial activity and overall biomass formation. The stabilization of the biomass and necromass formed, however, depends on the mineralogy and the redox dynamics of the soil. Minerals, in particular clay minerals, were shown to be the most correlating factors of necromass and SOM storage in various forest soils (Angst et al., 2018; Angst et al., 2021; Kleber et al., 2021).

The stabilization of non-living ¹³C-labelled microbial biomass by the soil matrix was tested in soils of two contrasting forest ecosystems (temperate forest in California; tropical forest in Puerto Rico (Throckmorton et al., 2015)). That particular study traced microbial biomass from fungi, actinobacteria, as well as Gram-positive and Gram-negative bacteria into soil density fractions. As expected, for both ecosystems the highest amount of soil C was found in the heaviest mineral-associated fraction. After prolonged incubation, the highest percentage of the microbial necromass-derived ¹³C was recovered in the mineral-associated fraction, irrespective of the type of microbial cell material. Mineral associations are thus more important for stabilizing microbial necromass in soil than the cellular origin. However, the exact fate of the necromass was under the strong control of site-specific edaphic factors. Fungal hyphae appear to have a relatively fast turnover, with a reported life span of only 7–10 days (Godbold et al., 2006). This would imply higher turnover times for fungal necromass compared to bacterial necromass, and a potential shift in the fungal/bacterial ratio when going from biomass to necromass (Liang et al., 2019). Other results showed a lower mineralization of ectomycorrhizal fungi (Schweigert et al., 2015).

Microbial necromass is stabilized like all other types of organic matter mostly by aggregation and bonding to the mineral matrix of the soil (Cotrufo et al., 2013; Angst et al., 2018; Angst et al., 2021; Kleber et al., 2021). It is generally considered that SOM is dominantly stabilized at the mineral phase as organo-mineral associated organic matter (MOM) (Kleber et al., 2015; Angst et al., 2018; Gao et al., 2019; Haddix et al., 2020; Angst et al., 2021; Gerrit et al., 2021). However, mineral association comprises a multitude of interactions dominantly considered as molecular interaction (Kleber et al., 2021). As microbes frequently live attached to particle surfaces, microbial necromass already is in direct contact to minerals when it is formed. In addition, due to the small size, microbial necromass can be associated with small-scale minerals, resulting in a considerable stabilization; however, these associations can also be metabolized by living microbes (Omoike and Chorover, 2006; Mikutta et al., 2007; Mikutta et al., 2011; Throckmorton et al., 2015; Gao et al., 2019; Kleber et al., 2021).

The mineral phase controls the long-term (up to millennia) C storage (Mikutta et al., 2007; Mikutta et al., 2009; Doetterl et al., 2015) as well as N storage in soil (Jilling et al., 2018). Biogeochemistry studies of MOM across long-term mineralogical soil gradients and chronosequences on the Hawaiian Islands (Mikutta et al., 2009) showed that SOM contents increased significantly with increasing surface area of poorly crystalline Fe and Al minerals. This correlation got lost, however, when formation of secondary minerals commenced. A significant impact of the plant cover of topsoils was detected, whereas mineral composition was more important in deeper layers. Studies in a glacier forefield chronosequence along a 120 ky ecosystem gradient (Mikutta et al., 2019) confirmed that MOM was formed through association with minerals of both microbial necromass and modified plant-derived compounds. Litter quality was less important for the development of MOM, and plant-derived C that had not been microbially processed was present in all soil depths.

The association with and occlusion in minerals and aggregates was considered to be the most important stabilization mechanism for microbial necromass in soil (Guggenberger et al., 1999) and Fe and Al oxides play a dominant role in these stabilization processes (Eusterhues et al., 2003; von Lützow et al., 2006; Schneider et al., 2010). Association of SOM to minerals is typically considered as interaction of separate molecules with mineral surfaces (Kleber et al., 2021) but often such materials are aggregations of biomolecules or cell envelope fragments of colloidal size in the nm range, which is not yet in the scientific focus. The association of microbial necromass with amorphous minerals or freshly precipitated metal oxides in soils has been demonstrated using various methods, suggesting important relations to the redox cycles of Fe (Ludwig et al., 2015; Kunhi Mouvenchery et al., 2016; Woche et al., 2017). The biogenic oxidation of reduced Fe, accompanied by precipitation, may result in incrustation of the cells and necromass materials, thus leading to strong preservation of cell organic matter on timescales up to 100,000 years under extreme conditions (Posth et al., 2014) as long as the Fe oxides are not reduced again. Microbially enhanced incrustation of SOM by Fe, Al, Mn, or Si oxides additionally slows down degradation (Kappler et al., 2005; Posth et al., 2014). Apart from sorption, the specific molecular architecture of the necromass fragments, as well as the self-aggregation of this material and the embedding in EPS of biofilm materials, contribute to physical separation of the materials, making them inaccessible for degrading enzymes. In addition to complexation with other organic compounds, microbial components and necromass are protected through occlusion within aggregates, creating physical barriers against enzymatic attack (Balesdent et al., 2000; Christensen, 2001; Rasmussen et al., 2005; Plaza et al., 2013). In addition, bacteria and necromass were found to be predominantly associated with clay particles forming clay “hutches” as a kind of shelter (Lünsdorf et al., 2000; Plaza et al., 2013).

Even if sorbed to minerals, microbial necromass is not protected from recycling by other microbes (Creamer et al., 2019), which is an important pathway for microbial resource mining. The authors identified two mechanisms of OM enrichment on two different minerals (feldspar and Al(OH)₃): the molecular enrichment of non-living OM and the enrichment of living or dead cells on surfaces. They also proved that microbial necromass is strongly sorbed to Al(OH)₃, making it poorly bioavailable. Feldspar in contrast has a lower accumulation potential for organic compounds in general. However, necromass retention on Al(OH)₃ was decreased in the presence of living bacteria compared to the control. This indicates that living cells attached to the minerals can recycle mineral-sorbed necromass C. Therefore, both mechanisms, stabilization by mineral-sorption as well as destabilization by mobilization, have to be considered. Mobilization of sorbed materials was particularly found for mineral-associated N-containing compounds (Turner et al., 2017; Creamer et al., 2019; Enggrob et al., 2020), supporting the microbial resource mining hypothesis.

Redox cycling strongly controls both C and P storage, e.g. in the rhizosphere of paddy soils (Huang et al., 2018; Wei et al., 2020; Kleber et al., 2021). The majority of stored P and C is associated with freshly deposited ferrihydrite, often formed by Fe-oxidizing bacteria in oxic microhabitats (Kappler et al., 2005; Posth et al., 2014). The current state of knowledge is that Fe-metabolizing bacteria are intimately linked to the Fe cycle in all environments and provide the key preservation of OM in soils and sediments (Posth et al., 2014). In addition, physical protection in concretions and aggregates is also correlated to changes in the moisture regime (Cates et al., 2019). Conversely, the release of C and P depends on Fe-reducing bacteria utilizing electron donors from root exudates as C and the energy source in anoxic habitats (Kappler et al., 2005; Achtenhagen, 2015). The dynamics of both processes depend on drying and rewetting cycles; there is also a strong link to N cycling. Excess of ammonium in the rhizosphere was shown to result in significant Fe-reduction-related ammonia oxidation (Feammox) accompanied by the release of soluble Fe²⁺ (Li et al., 2019a).

After exposure to redox fluctuations mimicking wet rice production with and without addition of rice straw, SOM and microbial community structures have been traced in soils with different mineralogy (Winkler et al., 2019). In comparison to continuously oxic conditions, lower overall mineralization was shown for the samples from redox cycling with a tendency to store more C from straw addition. Under cycling, more straw-derived C was allocated to C-retaining MOM, irrespective of the soil type (Winkler et al., 2019). Dissolution and precipitation of Fe oxides is essential for MOM formation and thus SOM protection, as more MOM was retained in soil with higher amounts of redox reactive minerals and cycling. This, however, requires a sufficient input of fresh OM as electron donor and C-source in order to induce sufficient microbial activity to induce Fe oxide dissolution (Posth et al., 2014). Similar effects of dissolution of amorphous Al(OH)₃ with subsequent co-precipitation with OM due to shifts in pH can also protect OM from decomposition (Mikutta et al., 2011), although the role of microbial activities for this process are not clear yet. Overall, mineralogy had no or only little effect on the turnover of straw-derived C. However, the capacity of the redox-active minerals is responsible for whether newly formed MOM replaces older MOM or is retained in addition to older MOM sorbed by the minerals (Winkler et al., 2019). We can hypothesize that these processes will apply in a similar way to necromass.

Redox cycling and the related dissolution and precipitation of Fe oxides are thus important key factors for SOM, in particular for necromass stabilization with continuous seasonal or annual cycling. In the wet period of wet-dry cycles, the redox potential of soil can drop dramatically, with SOM released from Fe oxides upon microbial reductive dissolution. With recurring oxidative conditions, fresh Fe oxides are formed that have low crystallinity and small particle sizes in the nm range. Furthermore, they cover existing surfaces, and thus offer large surface areas for sorption, protecting SOM and causing encrustation of larger biomaterials. Co-precipitation of oxidized Fe with necromass fragments (cell envelopes) was shown to result in strong stabilization through spatial protection (Achtenhagen, 2015). Such dynamics provide a close link between SOM stabilization and release of C, P and organic N, including microbial biomass building blocks.