Contrasting geochemical and fungal controls on decomposition of lignin and soil carbon at continental scale

Contrasting geochemical and fungal controls on decomposition of lignin and soil carbon at continental scale | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Contrasting geochemical and fungal controls on decomposition of lignin and soil carbon at continental scale Wenjuan Huang, Wenjuan Yu, Bo Yi, Erik Raman, Jihoon Yang, KE Hammel, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-2086399/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 19 Apr, 2023 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Lignin is an abundant and complex plant polymer that may limit litter decomposition, yet lignin is sometimes a minor constituent of soil organic carbon (SOC). Accounting for geographic diversity in soil characteristics might reconcile this apparent contradiction. We tracked decomposition of a lignin/litter mixture across North American mineral soils using lab and field incubations. Cumulative lignin decomposition varied 18-fold among soils and was strongly correlated with bulk litter decomposition, but not SOC decomposition. Legacy climate predicted decomposition even in the lab. Impacts of nitrogen availability were minor compared with geochemical and microbial properties, which had contradictory relationships with lignin and SOC decomposition. Lignin decomposition increased with some metals and fungi, whereas SOC decomposition decreased with all metals and was weakly related with fungi. Soil properties differentially impact decomposition of lignin and litter vs. SOC across broad geographic scales, linking short-term decomposition to differences in organic matter among ecosystems. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Lignin is one of the most abundant biopolymers in the terrestrial biosphere and protects other components of plant tissue from microbial attack. Lignin decomposition is critically linked to our understanding of the fate of plant litter, and it might also influence the composition and persistence of soil organic carbon (SOC). However, lignin’s importance in controlling litter and SOC decomposition remains controversial. It was traditionally thought that lignin limits overall litter decay after unprotected C compounds have been depleted 1 , 2 and that lignin derivatives are major constituents of SOC 3 , 4 . However, recent work indicated that lignin might decompose fastest during early stages of litter decomposition, perhaps as a consequence of co-metabolic degradation with labile C substrates 5 , and that availability of labile C may increase overall lignin loss 6 . Analyses of extractable lignin phenols suggested that lignin might even decompose faster than bulk SOC 7 , and a recent framework proposed that lignin-derived C is less persistent in soil than microbial-derived C 8 . The contradictory findings related to lignin decomposition might be related to biogeochemical differences among ecosystems. The persistence of lignin relative to other C substrates might vary systematically with geochemical and microbial characteristics across diverse soils 9 . This is because the controls on lignin decomposition are not necessarily the same as controls on bulk litter or SOC decomposition, yet these decomposition processes have rarely been investigated together. Climate and the ratio of lignin to nitrogen (N) can effectively predict litter decomposition at site to continental scales 2 , but these variables may have different relationships with lignin or SOC decomposition. For example, although high precipitation generally increases litter decomposition 2 , it can decrease decomposition of lignin or SOC by increasing mineral weathering and chemical stabilization of C with reactive metals 10 , 11 . Greater litter N content may increase lignin and litter decomposition by alleviating microbial N limitation 1 , 6 , whereas increased N availability may also decrease decomposition of lignin or SOC by suppressing the production of oxidative enzymes 4 . Both mechanisms may occur in the same soils depending on the stage of litter decomposition; N may stimulate early stages while inhibiting later stages of litter decay 6 , 12 . A consistent measurement of lignin decomposition across greatly differing soils would enhance our understanding of the roles of climate and N availability in regulating decomposition and their relative importance as compared to other soil properties. In addition to the traditional variables of climate and N availability, soil geochemical characteristics might also be important predictors of lignin and litter decomposition in ways that differ from bulk SOC. Soil minerals and metals can protect SOC from microbial decomposition through sorption, co-precipitation and polyvalent cation bridging 13 . In some soils, lignin-derived C may preferentially associate with Fe and Al relative to bulk litter or bulk SOC 14 – 16 . Correspondingly, reactive forms of Fe and Al might be more important for limiting decomposition of lignin vs. other compounds in litter or bulk SOC. Similar to Fe and Al, Mn might protect C by physical or chemical mechanisms 17 , yet because some forms of Mn are powerful oxidants, increased Mn availability can also stimulate lignin decomposition 18 . Calcium might also have a contrasting relationship between decomposition of different substrates: Ca promotes physicochemical protection of SOC and often negatively correlates with SOC decomposition 19 , yet Ca availability may stimulate lignin-degrading fungi and correlate positively with lignin and litter decomposition 20 . However, the consistency of relationships between metals and decomposition rates of lignin, litter, and/or SOC across diverse ecosystem types remains unresolved. Besides geochemistry, the composition and abundance of soil microbes may also impact decomposition. Many microbial taxa may perform similar functions, and it remains elusive whether soil microbial composition explains variation in process rates 21 . Yet, microbial composition might differentially impact the decomposition of different C forms. On one hand, due to lignin’s complex structure of heterogeneous aromatic subunits linked by non-hydrolyzable bonds, only a small subset of microbial taxa, primarily white-, brown-, and soft-rot fungi, and certain bacteria, has been conclusively demonstrated to cleave the lignin macromolecule at the propyl side chain, which is likely the rate-limiting step in lignin decomposition 22 – 24 . On the other hand, microbial community composition may be less important for decomposition of SOC than for lignin simply because physical restrictions on microbial access to C substrates may predominantly limit SOC mineralization 25 . However, few studies have directly compared the relationships of fungi with decomposition of lignin and litter vs. SOC, and those that did mainly quantified rates of lignin decay along with individual fungal taxa at local scales 26 . Fungal community analyses paired with measurements of lignin, litter, and SOC decomposition across a set of heterogenous soils could help us to evaluate how composition, diversity, and abundance of soil fungal communities are linked to decomposition among ecosystems. Here, we measured decomposition of lignin, bulk litter, and SOC via a uniform and quantitative isotopic method from mineral soil samples collected across broad biophysical gradients to test biogeochemical controls on decomposition of these substrates. We used 20 sites from the National Ecological Observatory Network (NEON) that span diverse ecosystems, climatic zones (tundra to tropics), and soil characteristics across North America (Fig. 1 a–c, Supplementary Table S1 and Fig. S1). Previous examinations of lignin decomposition have often relied on semi-quantitative or indirect methods, such as acid-unhydrolyzable residue to approximate lignin content 27 , oxidation of simple substrates as a measure of potential ligninolytic enzymes 4 , or use of lignin monomers rather than polymers in incubation experiments 28 . These methods can substantially underestimate or overestimate the lignin content of litter and soil as well as the activities of ligninolytic enzymes 29 , 30 . However, C isotope-labeled, high-molecular-weight synthetic lignins provide a tracer that allows unambiguous quantitative measurement of lignin decomposition 31 . We combined isotope-labeled lignin with natural-abundance litter derived from a C 4 grass and added these mixtures to separate soil samples for incubation in the lab, enabling us to quantify lignin, litter and SOC decomposition over time (Fig. 1 d) 4 , and to assess their relationships with climatic, N-related, geochemical and microbial factors across soils. Lignin decomposition was also measured in field-incubated samples. We hypothesized that (1) lignin decomposition predictably varies with soil geochemical characteristics (reactive minerals and metals) and fungal communities at the continental scale, in addition to climatic and N-related variables; and (2) the predictors for lignin decomposition are similar with litter decomposition, while these predictors have different relationships with SOC decomposition due to specific interactions of lignin with metals and microbial communities. Results Decomposition rates of lignin, litter, and SOC. The temporal dynamics of lignin C decomposition were generally more variable than litter C or SOC decomposition, although instantaneous lignin decomposition was about 4-fold lower than litter and soil decomposition, on average, when normalized by C mass in these pools (Fig. 2 ). Lignin decomposition rate generally decreased over time; however, in some soils it transiently increased over timescales of months or was still increasing at the end of incubation (after 18 months) (Fig. 2 a and Supplementary Fig. S2). Litter decomposition rate also increased transiently over time in some sites (Fig. 2 c), and it was significantly related to instantaneous lignin decomposition rate as indicated by Pearson correlation ( P < 0.01). The temporal pattern of lignin C decomposition differed from SOC decomposition, which generally showed a declining trend at the end of the incubation (Fig. 2 e); instantaneous decomposition of lignin C and SOC were not statistically related ( P > 0.05). At the end of the lab incubation (571 d), cumulative C decomposition relative to initial C was 1.7–31.4% for lignin, 2.0–53.0% for litter and 6.3–91.2% for SOC (Fig. 2 b, d, and f). The cumulative decomposition of lignin and SOC was similar between 0–15 cm and 15–30 cm soil samples, while cumulative litter decomposition was significantly lower in the deeper soil (34% for 0–15 cm vs. 23% for 15–30 cm; P < 0.01). We also conducted a separate N addition experiment to test the effects of N availability on lignin and litter decomposition in the 0–15 cm soils. Lignin and litter decomposition rates were slightly increased by N addition in most sites throughout the lab incubation (Supplementary Fig. S3). Overall, cumulative decomposition of lignin and litter was slightly but significantly increased by N addition after the 571-d incubation ( P < 0.01 for both; Fig. 3 ). On average, the N addition treatment increased cumulative lignin decomposition by 1.6% and litter decomposition by 6.2% across the 20 sites. Neither lignin nor litter decomposition was significantly depressed by N addition at any site after 18 months (Fig. 3 ). We further compared lignin decomposition measured in the lab with field incubations of 0–15 cm soils. Cumulative lignin C decomposition measured in the field-incubated samples after approximately 1 y showed higher variation among samples and overall higher rates than observed in the lab (Supplementary Fig. S4), corresponding with overall higher fungal quantity in the field than in the lab (Supplementary Fig. S1). Total lignin C loss relative to initial lignin C concentrations averaged 9–63% across the 20 sites in the field, while the site-averaged lignin C loss in the lab ranged from 3–16% (Supplementary Fig. S4). There was no significant correlation between field and lab lignin C loss. Variation in biogeochemical predictors among soils. Along with climatic factors (mean annual temperature, MAT, and mean annual precipitation, MAP), we selected 25 biogeochemical predictors and separated them into three categories, including those related to N availability (11), geochemistry (8), and microbes (6) (Supplementary Table S2). The variation of these biogeochemical predictors was very high among sites and was generally similar between the lab and field incubation datasets (Supplementary Fig. S1). For the N predictors, total soil N was 0.1–32 mg g − 1 and C/N was 8–58. Both NH 4 + -N (0–104 µg N g − 1 ) and NO 3 − -N (0–937 µg N g − 1 ) tended to increase with incubation time in the lab (Supplementary Fig. S1a). For the geochemical predictors, soil pH was 4.0–9.2, and soil particle size (silt + clay, 5–91%) and metals had at least one order of magnitude difference among sites (0–30 mg g − 1 Al ox ; 0–20 mg g − 1 Fe HCl , 0–78 mg g − 1 Fe ox ; 0–54 mg g − 1 Fe cd−ox , 0.0–3.3 mg g − 1 Mn cd , 0–55 mg g − 1 Ca cd ; Supplementary Fig. S1b). Microbial community variables exhibited as much as four orders of magnitude difference among sites, with fungal quantity of 1.1×10 5 –2.6×10 9 gene copies g − 1 , bacterial quantity of 5.2×10 9 –9.8×10 11 gene copies g − 1 , fungal-to-bacterial ratio of 5.8×10 − 6 –6.2×10 − 2 , and a fungal diversity index of -39–73 (i.e., the residual of the Chao1 index for ITS ASV’s, Supplementary Fig. S1c). Microbial quantity changed with time: fungal quantity increased in the 9-month incubated samples vs. initial soil samples, and fungal quantity further increased and bacterial quantity also increased after 14 months vs. 9 months (Supplementary Fig. S5a). Relationships of C decomposition with biogeochemical predictors. Cumulative decomposition of lignin and litter at the end of lab incubation tended to have similar relationships with biogeochemical predictors, whereas lab lignin and SOC decomposition had generally opposite pairwise relationships with those same biogeochemical predictors (Fig. 4 ). We assessed these relationships after 6, 12 and 18 months of cumulative decomposition (Supplementary Figs. S6 and S7), and they changed little over time, such that we focused our subsequent analysis on the 18-month (571 d) dataset. Different axes of fungal community composition correlated with lignin and litter vs. SOC decomposition. MAP was significantly and positively related to lignin decomposition but negatively related to SOC decomposition ( P < 0.01). Lignin decomposition had a significantly negative relationship with NO 3 − -N after 1 m of incubation ( P < 0.01), while litter decomposition was positively related to total N, NH 4 + -N, NO 3 − -N and inorganic N (NH 4 + -N + NO3 − -N) after 18 months ( P < 0.05 for NH 4 + -N and P < 0.01 for the other variables). Soil pH was significantly and negatively related with lignin decomposition but positively related with SOC decomposition ( P < 0.01). The silt + clay content and some soil metals (Fe cd−ox and Mn cd ) had significant and positive relationships with lignin decomposition but negative relationships with SOC decomposition ( P < 0.01). Fungal quantity had a positive relationship with lignin and litter decomposition but a negative relationship with SOC decomposition. Compared with lab lignin decomposition, field lignin decomposition was also closely related to some predictors within climatic, N, geochemical and microbial categories, although not all relationships were consistent. Specifically, fungal quantity had a positive relationship with field lignin decomposition, consistent with the lab lignin decomposition. However, the field lignin decomposition had a negative relationship with MAP and a positive relationship with soil pH, which was inconsistent with the lab lignin decomposition. Importance of biogeochemical predictors for C decomposition. We used two statistical model approaches (linear mixed model, LMM and random forest model, RFM) to identify the most important soil geochemical, microbial, N, and climatic predictors of decomposition (Fig. 5 ). The RFM partial dependence plots showed that many relationships between predictors and response variables were approximately linear until predictors increased above the 90th or larger percentiles, where response variables became approximately constant (Supplementary Figs. S8 to S11). The LMM showed that soil pH and fungal composition were the strongest predictors for lab lignin decomposition, and that MAT, fungal quantity, Mn cd , Ca cd , silt + clay, Fe ox , and soil C/N could also improve the final model (Fig. 5 ). Soil pH and Fe ox had negative relationships with lab lignin decomposition while all other predictors had positive relationships. These predictors explained 43% of the observed variance in lab lignin decomposition; the overall model (including random effects for site) explained 45%. A slightly higher proportion of variation in lab lignin decomposition (R 2 = 0.51) was explained in the RFM. The optimal RFM included the same predictors as the LMM (except for Fe ox ), with the addition of bacterial quantity, MAP, and Fe cd−ox . Predicted lignin decomposition decreased with pH and increased with most other predictors (Supplementary Fig. S8). However, predicted lignin decomposition varied nonlinearly with soil C/N, with a minimum at C/N = 20. Ten fungal genera occurring in more than 10 samples had significant ( P < 0.05) correlations with lignin decomposition (Supplementary Table S3): positive for Pleotrichocladium, Geomyces, Trichocladium, Mycena, Hypochnicium, Solicoccozyma , and Mortierella and negative for Cladophialophora, Pseudofabraea , and Glarea . Lignin decomposition in the field and lab shared many of the same predictors in the LMMs, including MAT, Fe ox , Ca cd , and Mn cd (Fig. 5 ). Fe ox showed a negative relationship while the other three predictors showed positive relationships with field lignin decomposition. These four predictors, along with MAP, soil C/N, soil N, pH, and Al ox , explained 31% of the variation in field lignin decomposition; the overall model with random effects explained 53% of the variation. All nine predictors also explained 39% of the variation in field lignin decomposition in the RFM. As observed in the LMM, predicted field lignin decomposition generally increased with increasing Mn cd and Ca cd and decreased with increasing Fe ox and Al ox in the RFM (Supplementary Fig. S9). Unlike lab lignin decomposition, predicted field lignin decomposition showed more complex relationships with MAT with a null relationship below 15°C and a positive relationship at higher temperatures, predicted lignin decomposition decreased with increasing MAP, and Fe cd−ox was not an important predictor. Similar to the lab lignin decomposition, predicted field lignin decomposition varied nonlinearly with soil C/N, with a minimum at a slightly lower C/N value of 13 (Fig. 5 ). Predictors of litter decomposition were generally similar to lignin decomposition. The LMM showed that Fe ox was the strongest predictor of litter decomposition, followed by fungal composition, Mn cd , Fe HCl , soil N, fungal quantity, bacterial quantity, MAT and pH (Fig. 5 ). Fe ox and pH were negatively while all other predictors were positively related to litter decomposition. These predictors explained 49% of the variation in litter decomposition, and the overall model with random effects explained 52%. The optimal RFM explained 60% of the variation in litter decomposition and included the same predictors as the LMM except for silt + clay and soil C/N. Fe ox , Al ox , and pH had weak negative relationships with litter decomposition in the RFM while most other predictors had positive relationships (Supplementary Fig. S10). Contrary to lignin and litter decomposition, microbial predictors were not important predictors of SOC decomposition in the statistical models (Fig. 5 ). The LMM showed that soil C/N, MAT, and silt + clay were the strongest predictors for SOC decomposition, and that MAP, Ca cd , Mn cd , Al ox , and pH were also important (Fig. 5 ). MAT and pH were positively while all other variables were negatively related to SOC decomposition. The predictors collectively explained 43% of the variation in SOC decomposition and the overall LMM (including random effects) explained 71%. The optimal RFM explained 55% of the variation in SOC decomposition, and included soil C/N and all geochemical and climatic predictors. Predicted SOC decomposition generally decreased with soil C/N, reactive minerals and metals, and MAP, and increased with MAT (Supplementary Fig. S11). Discussion Our continental-scale data demonstrated that lignin decomposition was not universally slow or fast when compared with decomposition of litter and SOC, but rather, it varied predictably among sites along with biogeochemical variables (Fig. 4 ). Consistent with our first hypothesis, a large proportion of variation in lignin decomposition could be predicted by soil geochemical and microbial properties, in addition to the traditional variables of climate and N availability (Fig. 5 ). Lignin decomposition was highly correlated with litter decomposition and these processes shared the same biogeochemical predictors, but they were generally different from predictors of SOC decomposition (Figs. 4 and 5 ), supporting our second hypothesis. Several soil geochemical factors had negative (Fe ox and Al ox ) and positive (Mn cd and Ca cd ) relationships with lignin and/or litter decomposition, while there were not any positive relationships between extractable soil metals and SOC decomposition (Figs. 4 and 5 ). Intriguingly, microbial variables including fungal composition and fungal and bacterial quantity were needed to explain variation in the decomposition of lignin and litter, but not SOC (Fig. 5 ). A relatively high amount of N addition stimulated decomposition of lignin and litter to a minor degree in most sites after 18 months (Fig. 3 ), but its overall impact was small when considering the wide range of decomposition across samples (Fig. 2 ). Inorganic N appeared to be less important than geochemical and microbial properties (Fig. 5 ). Long-term climatic predictors (MAT and MAP) could explain variation in both field and lab decomposition, but the actual vs. legacy climate, as reflected by the field vs. lab experiments, respectively, had different relationships with decomposition (Fig. 5 ). Overall, our data showed the particular importance of geochemical and microbial predictors for lignin and litter decomposition, and their differing relationships with SOC decomposition, which collectively supported different aspects of classic and modern views of decomposition. The strong correlation between lignin and litter decomposition and the similar biogeochemical predictors of these processes support the classic view that lignin decomposition is tightly coupled with overall litter decomposition 1 , 27 . However, we found that decomposition of SOC was unrelated to decomposition of lignin and litter, and that these processes often had contrasting relationships with biogeochemical predictors. This finding is inconsistent with the classic idea that the slow decomposition of lignin residues is an important factor that limits decomposition of total SOC 3 . The disparate rates and predictors of lignin and SOC decomposition support the modern notion that lignin depolymerization is not necessarily a primary bottleneck for SOC decomposition 7 , 8 . Furthermore, our dataset provides a mechanistic explanation for the decoupling of lignin and SOC decomposition by highlighting their differing relationships with geochemical and microbial variables. Differing geochemical predictors for decomposition of lignin and litter vs. SOC. We found that geochemical variables often had different relationships with lignin and litter decomposition than with SOC decomposition (Figs. 4 and 5 ). Some geochemical variables had negative (e.g., Fe ox and Al ox ) and positive (e.g., Fe HCl , Mn cd and Ca cd ) relationships with lignin and/or litter decomposition, while others (e.g., Fe HCl , Fe ox , Fe cd−ox , Al ox , Mn cd , and Ca cd ) mainly had negative correlations with SOC decomposition. The metals extracted from these NEON soils likely represent ions (e.g. Ca cd ), metals dissolved from specific mineral phases of varying crystallinity (e.g., Fe HCl , Fe ox , Fe cd−ox ), or a mixture of ions and mineral phases (e.g., Mn cd and Al ox ) 32 . Soil mineral and metal cations as well as fine particles (silt + clay) are important predictors of SOC concentration due to protection by sorption, precipitation, and aggregation 13 , 19 , 33 . However, such SOC stabilization mechanisms may not necessarily correspond with the short-term decomposition of newly added C sources. Protective effects of soil metals and minerals were mainly applicable for SOC decomposition, and conversely, some of these same variables were actually associated with greater decomposition of lignin and litter. We found positive associations of some soil metals (e.g., Fe cd−ox , Mn cd and Ca cd ) with lignin and litter decomposition (Figs. 4 and 5 ), consistent with catalytic or biological roles of soil metals for organic matter decomposition demonstrated in other studies 18 , 20 , 34 . The finding that Fe cd−ox was positively related and Fe ox was negatively related with lignin and litter decomposition was consistent with multiple functional roles of Fe, which might stimulate decomposition or provide protection depending on C molecular composition and/or redox environment 16 , 34 . Moreover, in samples with greater Mn and Ca, lignin and litter decomposition increased whereas SOC decomposition decreased (Figs. 4 and 5 ). Our results were consistent with the importance of Mn-promoted degradation and protection of organic C 17 , depending on particular C forms. Besides the protective effects of Mn on SOC, Mn also can promote lignin decomposition by stimulating the activities of lignin-degrading enzymes and oxidizing lignin via redox cycling 35 , and Mn-stimulated lignin decomposition likely increased overall litter decomposition 18 . Whereas the negative correlation of SOC decomposition with Ca was likely due to protective cation bridging with soil minerals 19 , the strong positive relationships between Ca and decomposition of lignin and litter agreed with previous studies showing that Ca was positively related to the extent of litter mass loss, and in particular, lignin degradation 20 , 36 . Ca is an essential component of the fungal cell wall and can increase the growth of white rot fungi 37 . We also found an overall positive relationship of silt + clay with lignin and litter decomposition, which might reflect multiple biological and physical factors that co-vary with particle size, as well as the potential for minerals to catalyze OM decomposition 38 . Overall, the role of certain metals in stimulating lignin and litter decomposition while suppressing SOC decomposition provides an explanation for fact that these processes may be coupled or decoupled to varying degrees 5 , depending on soil characteristics. Intriguingly, field lignin and lab lignin decomposition had opposite pairwise relationships with several biogeochemical predictors, such that field lignin and lab SOC decomposition also had several biogeochemical predictors in common (Fig. 4 ). However, most of these relationships disappeared in the statistical models after other predictors were accounted for (Fig. 5 ), and they may have been due to wider ranges of predictor and response variables in the field samples (such as fungal abundance and Fe ox ), and the strikingly different climate conditions in the field than in the lab (Supplementary Figs. S1 and S4). Microbial predictors explained decomposition of lignin and litter but not SOC. Consistent with our hypotheses, composition and quantity of overall fungal communities explained variation in lignin and litter decomposition (Figs. 4 and 5 ). However, the composition of known “rot” fungi was insignificantly ( P > 0.05) correlated with lignin decomposition and summed relative abundance of these fungi had a weak but positive relationship ( P = 0.02; data not shown). Seven genera occurring in > 10 samples were significantly ( P < 0.05) and positively correlated with lignin decomposition (Supplementary Table S3), but only three have been reported to possibly degrade lignocellulose: Trichocladium (soft rot), Mycena , and Hypochnicium (white rot) 39 – 41 . These results indicate that the most commonly studied lignin-degrading fungi (i.e., the known “rot” fungi) were not necessarily the most important lignin-degrading organisms in our continental-scale dataset 24 . Consistent with a previous study across North America 42 , we found that fungal communities were highly heterogenous across sites and even within plots; e.g., only 65 of 342 fungal species occurred in > 10 samples. This finding further suggests that specific fungal taxa possibly responsible for lignin decomposition varied with locations and even depths at the same plot. Bacterial quantity was also related to both lignin and litter decomposition (Figs. 4 and 5 ). Although bacteria may degrade lignin directly 22 , they might simply be responding to increased C availability as a consequence of fungal lignin decomposition, or may synergistically interact with fungi to promote lignin decomposition 43 . Previous work suggested that microbial community composition does not necessarily influence ecosystem processes because of high functional redundancy 21 , 44 . Yet our findings, in line with previous studies 45 , 46 , clearly showed that differences in fungal community composition and abundance among diverse soils explained variation in lignin and litter decomposition rates when temperature and moisture were held constant. And furthermore, the differing relationships between fungal composition and decomposition of lignin and SOC provide another mechanistic explanation for the observed decoupling of these processes. Contrary to lignin and litter decomposition, microbial variables were less important predictors of SOC decomposition (Fig. 5 ), despite high variation in fungal composition and richness and bacterial and fungal quantities across soils (Supplementary Figs. S1 and S5). Different axes of fungal community composition correlated with decomposition of lignin and litter (PC1) vs. SOC (PC2; Fig. 4 ). The importance of soil microbial communities in driving SOC decomposition remains poorly understood and contrasting findings have been reported. On one hand, significant relationships among SOC decomposition rate and microbial community composition, biomass, and richness have been reported 47 . On the other hand, microbial biomass, community structure or specific activity were weakly related to SOC decomposition 48 . Consistent with the latter findings, we found that despite their significant pairwise correlations (Fig. 4 ), microbial predictors including fungal community composition, fungal quantity, and bacterial quantity were not related to SOC decomposition after accounting for other variables (Fig. 5 ). This finding supports the hypothesized yet not broadly demonstrated view that SOC turnover is dominantly determined by decomposer access to SOC 25 . Because of the spatially heterogenous soil environment, a large proportion of SOC is probably stored in small pores 49 . The added lignin and litter were likely not protected in those pores, because they were gently mixed into the soil and the particles were much larger than the prevailing pore sizes. This might explain why although C substrates were mixed into soil in this experiment, their decomposition was still measurably related to fungal community composition and quantity. Interestingly, MAT and MAP of the study sites explained variation in organic matter decomposition (Figs. 4 and 5 ) even under the common conditions of temperature and moisture imposed in the lab incubation. This is consistent with previous findings that climate history was an important determinant of litter and SOC decomposition, possibly by shaping functional responses of decomposer communities and/or via correlation with soil minerals 11 , 50 , 51 . Decomposer communities could adapt to temperature and moisture regimes via changes in enzyme properties and community composition 52 . Climate greatly impacts soil weathering 10 , 11 , which results in soil geochemical changes that might directly influence OM decomposition, as suggested by our data (Figs. 4 and 5 ). Because the climate variables used as predictors reflected either the actual differences in climate during the field decomposition or the legacies of prior differences in climate during the lab decomposition, respectively, it was perhaps not surprising that MAP had different relationships with lab and field lignin decomposition (Supplementary Figs. S8 and S9). Nevertheless, all organic matter decomposition variables were positively related to MAT in the statistical models (Fig. 5 ), suggesting that the legacy effect of soil MAT was stronger than the legacy effect of MAP on OM decomposition. Small stimulation of litter and lignin decomposition from increased N availability. The impact of N on decomposition remains controversial. Whereas greater N availability may increase lignin and litter decomposition rates by alleviating microbial N limitation, it might also decrease late-term decomposition by suppressing oxidative enzymes involved in lignin breakdown 1 , 4 . Despite a negative relationship between initial nitrate and lab lignin decomposition (Fig. 4 ), on balance inorganic N addition led to a small net stimulation of lignin and litter decomposition after 18 months (Fig. 3 ), and frequently higher decomposition in the N addition vs. control treatments over time (Supplementary Fig. S3). The positive response of lignin and litter decomposition to N addition might imply that microbial growth was N-limited in many sites. Nonlinear relationships between soil C/N and lignin decomposition might reflect multiple mechanisms between N and lignin decomposition (Supplementary Figs. S8 and S9). The positive relationships between total soil N and litter decomposition (Fig. 4 ) may be related to greater N availability which could alleviate microbial N limitation and thus facilitate decomposition. The negative relationship between total N and SOC decomposition (Fig. 4 ) might result from reduced microbial mining of SOC because of increasing N availability. Overall, despite the stimulation of lignin and litter decomposition, the effects of N on lignin and litter decomposition were relatively small in comparison with variation across sites, even following a substantial addition of inorganic N (Figs. 2 and 3 ). Therefore, when considering continental-scale variation in biogeochemical properties, variation in N availability may be a less important driver of decomposition than sometimes assumed. Implications for geochemical and microbial control of SOC concentration and composition. Comparison of our results with other recent observations from NEON soils indicates that the differing controls on decomposition of lignin and litter vs. SOC implied by our data may contribute to variation in SOC concentration and organic matter composition among ecosystems. Many of the same variables that predicted C decomposition in the lab incubation also predicted differences in SOC concentration and the distribution of SOC between SOM size fractions, defined as chemically dispersed particulate organic matter (POM, > 53 µm) and mineral-associated organic matter (MAOM, < 53 µm), which were described in previous studies of NEON soils 33 , 53 . Silt- and clay-sized minerals and reactive Fe phases in particular have long been thought to protect SOC from decomposition, even though the relationships among these variables can be relatively weak across large datasets 33 , 54 . Here, we found that the magnitude or even the sign of the pairwise correlation or model coefficient between decomposition and silt + clay or Fe in various extractions (Fe HCl , Fe cd−ox , Fe ox ) often differed between lignin/litter and SOC (Figs. 4 and 5 ). These differences could influence SOM composition while explaining the context-dependency of relationships between Fe and SOC concentration in other datasets 33 , 54 . For example, negative relationships of Fe ox with lignin and litter decomposition (Fig. 5 ) could help explain the positive relationship between Fe ox and the increasing proportion of SOC in POM vs. MAOM, which we observed in our previous study with the same soils 53 . That is, Fe ox could increase POM by disproportionately decreasing rates of lignin decomposition relative to bulk SOC, consistent with the view that POM is mostly composed of decomposing plant detritus which may aggregate with metals 55 . Similarly, the positive relationship between silt + clay and lignin decomposition and its negative relationship with SOC decomposition is consistent with our previous finding that increased silt + clay was associated with lower SOC in POM vs. MAOM 53 . The decrease in POM vs. MAOM with increasing silt + clay might simply be due to increased capacity for mineral protection, but it might also be linked to increased catalysis of lignin decomposition by metals and/or minerals in these fine particle fractions 38 ; Mn and Ca were both associated with greater decomposition of lignin, but not SOC (Fig. 5 ). Together, the contrasting relationships of silt + clay and Fe ox with decomposition of lignin and litter vs. SOC provide an explanation for why these variables may be poor predictors of SOC concentration over broad scales, even though they may be related to the physical forms (POM vs. MAOM) of SOC. In summary, using a quantitative isotopic method, we found that decomposition of lignin varied 18-fold among soils sampled from sites across North America and incubated in a common environment. Lignin decomposition was always slower than but was strongly related to bulk litter decomposition. Differences in lignin decomposition among sites were strongly related to biogeochemical predictors, in a manner that was similar to bulk litter decomposition but differed from SOC decomposition. Different axes of fungal community composition were related to decomposition of lignin and litter compared to SOC, and metals often positively correlated with lignin decomposition even though they had a neutral or negative correlation with SOC decomposition. Similarities in controls on lignin vs. bulk litter decomposition reinforces the traditional view that lignin is tightly coupled with overall litter decay over timescales of months to years. In contrast, the difference in controls on lignin and litter decomposition vs. SOC supports the modern notion that lignin depolymerization is not a primary bottleneck for SOC decomposition. While substantial research has focused on N dynamics as controls on litter decomposition, our data showed that while significant, the influence of N availability on decomposition of lignin and litter mixed into mineral soils was often smaller than other geochemical and microbial factors. Legacies of previous climates may predict decomposition rates under similar conditions of temperature and moisture. Our data suggest the critical need for mechanistic models to account for contrasting geochemical and microbial controls on decomposition of lignin and litter vs. SOC, in addition to the traditional variables of climate, residue quality, and nutrient availability. Methods Experimental design. We used 20 sites from National Ecological Observatory Network (NEON) to examine decomposition of lignin, bulk litter, and SOC and to test biogeochemical (geochemical, microbial) controls on decomposition of these substrates, in addition to N-related and climatic variables. Soils amended with C stable isotope ( 13 C) labeled and un-labeled lignins and a single natural litter source were incubated in the lab to quantify lignin, litter and SOC decomposition over 18 months (Fig. 1 d). An additional incubation experiment was also conducted to test the effects of N addition on lignin and litter decomposition. The results of lignin decomposition and its predictors from the lab incubation were further compared with those from a field incubation. Site selection and soil sampling. NEON is a U.S. based, continental-scale ecological monitoring network that provides open data, samples, and research infrastructure to reveal how ecosystems are responding to environmental change 56 . For this project, we selected 20 NEON terrestrial sites, denoted by their acronyms as follows: BONA, CPER, DSNY, GRSM, HARV, KONZ, LENO, NIWO, ONAQ, OSBS, PUUM, SJER, SRER, SCBI, TALL, TOOL, UNDE, WREF, WOOD, YELL (Fig. 1 a). These sites span wide edaphic, climatic and ecosystem gradients to maximize differences in soil geochemical and microbial communities (Supplementary Fig. S1). They encompass 9 out of the 12 soil orders in the United States Department of Agriculture (USDA) soil classification system (no Histosols, Oxisols, or Vertisols; Supplementary Table S1). The sites had mean annual temperature (MAT) of -9–22 o C and received 262–2657 mm of mean annual precipitation (MAP) 57 . The sites included diverse ecosystem types, such as tundra, forest, wetland, grassland, shrubland and desert. Soils at each site were sampled by NEON staff during the growing season of 2019 (April–August; later sampling occurred at Alaska sites where soils did not thaw until July or August). Mineral soil samples were collected at two depths (0–15 cm and 15–30 cm), after removing any surface litter or organic horizon (Supplementary Table S1), using a 2- to 5-cm diameter corer, according to the standard NEON sampling procedure for that particular site. At each site, samples were collected around the perimeter of one 40 × 40-m “distributed base plot” which was selected to represent the dominant upland vegetation type and soil type of that site, whenever possible, in accordance with site access constraints. Soil at the KONZ site was collected only at 0–15 cm due to the shallow soil depth. Each plot had 16 replicates (n = 16), denoted sampling points 1–16 hereafter. Point 1 was located 4 m west and 4 m south from the SW corner of each plot, and the other points were located in counterclockwise sequence at 12-m intervals around the perimeter of the plot, each located 4 m outside of the plot boundary (Fig. 1 b). Soil cores from each point were collected and shipped overnight on ice (~ 4 o C) to Iowa State University (ISU) for use in laboratory and field incubations. Soil from each sample was gently homogenized inside a plastic bag after any coarse roots, macrofauna, or rocks were manually removed. We did not sieve samples except for ONAQ and SRER, where rocks were abundant and were removed by sieving soil through a 2-mm sieve. Lab incubation experiments. Soils from four sampling points at two depths per site were used for lab incubation and biogeochemical analyses, totaling 156 samples. The four sampling points were mainly selected at the odd number in the middle of each side of the 40 x 40-m plot (red circles in Fig. 1 b), although sampling points from some sites were selected at the numbers next to the middle numbers if soils were not available for both layers. Subsamples of soils used for the lab incubation experiment were brought to field moisture capacity, which was determined for each soil by saturating an additional 20–30-g subsample placed on filter paper in a funnel, and then measuring gravimetric water content following 48 h of drainage. Subsamples (1 g dry mass equivalent) from each sampling point and depth were incubated under each of three separate substrate treatments to partition C decomposition among three sources, using measurements of d 13 C values of CO 2 . We quantified decomposition of C from extant soil organic matter, C from added litter (senesced leaves of Andropogon gerardi , a C 4 grass), and a specific C atom (the C β position of the propyl sidechain) in lignin that was precipitated on the added litter. The lignin was prepared as described previously 5 . The detailed method is described in the SI. Substrate treatments were: (1) soils alone (control); (2) soils amended with A. gerardi litter precipitated with trace natural abundance 13 C lignin (soil + litter + unlabeled lignin); (3) soils amended with A. gerardi litter precipitated with trace lignin labeled with 99 atom percent 13 C at the C β position of each lignin C 9 substructure (soil + litter + 13 C β -labeled lignin). We added uniform litter and synthetic lignin to each of the mineral soil to focus on soil biogeochemical gradients rather than substrate quality. Soils were gently mixed with the litter + lignin mixture in a 250:25:1 ratio of soil:litter:lignin (1 g dry soil mass equivalent was mixed with 100 mg litter and 4 mg lignin). To prepare the litter + lignin mixture, the unlabeled or labeled lignins were precipitated in a 1:25 mass ratio on dried and finely ground leaf litter of A. gerardi (41.9% C, 0.41% N, and δ 13 C = -12.6‰; see SI for more details). The 20 NEON sites comprise ecosystems ranging from C 3 -dominated forest and grassland sites to mixed C 3 –C 4 grasslands and/or plants with Crassulacean acid metabolism, such that the δ 13 C value of the added C 4 litter was always more positive than δ 13 C value of CO 2 derived from soil organic matter at a given site. Soil samples were incubated under oxic conditions in the dark at 23 o C for 571 d. Soil was kept in an open 50 mL centrifuge tube inside a glass jar (946 mL) sealed with a gas-tight aluminum lid with butyl septa for headspace gas purging and sampling. The jars were flushed with CO 2 -free air following periodic headspace sampling as described below, and CO 2 concentrations remained below 5000 ppm during the incubation. Soil moisture was monitored by recording the mass of each sample, and water was added every month before 179 d and every other month thereafter (due to the less frequent gas sampling) to replenish vapor lost during headspace flushing. To monitor instantaneous decomposition over time and to avoid CO 2 saturation in the jar, headspace gas was initially measured at 4 d and 11 d, every other week for another 140 d, and then every month after 179 d (for a total duration of 571 d). The CO 2 concentrations and their δ 13 C values were measured by a tunable diode laser absorption spectrometer (TDLAS, TGA200A, Campbell Scientific, Logan, UT) immediately prior to flushing the headspace 58 . Because jars remained sealed between headspace sampling events, we were able to quantify the entire cumulative production of CO 2 and its δ 13 C value from each replicate over the course of the experiment. The CO 2 production from soil was measured on samples with no addition of litter and lignin, and CO 2 from litter and 13 C β -labeled lignin was calculated by two-source mixing models that used measurements from the litter + unlabeled lignin and litter + 13 C β -labeled lignin treatments, respectively 16 , 58 (see SI for more details). The C decomposition from soil, litter and lignin were expressed as percentages of their initial C masses (41.9 mg for litter and 264 µg for the 13 C β atom of the labeled lignin, and a variable amount for SOC; Supplementary Table S1). We also conducted an N addition experiment to test the effects of N availability on lignin and litter decomposition, using additional subsamples of the 0–15 cm soils collected from the four sampling points described above. For this experiment, the subsamples amended with litter + unlabeled lignin or litter + 13 C β -labeled lignin were also amended with NH 4 NO 3 at 50 mg N g − 1 . Note that the amount of N that we added is relatively high but comparable to the level of inorganic N often observed in agricultural fields after fertilization 59 . Briefly, 51 mL of 0.0386 mol L − 1 NH 4 NO 3 was added to soil samples, and then more water was added as necessary to achieve field moisture capacity. Sample incubation and gas measurements were the same as described above and conducted over 18 months. Field incubation experiments. The 0–15 cm soils from all 16 sampling points at each site were used for field incubation (Fig. 1 b). Soil subsamples (4.5 g dry mass equivalent) were gently mixed with litter + unlabeled lignin or litter + 13 C β -labeled lignin according to the mass ratios and substrate treatments described above. The soil + litter mixtures were then transferred to mesh bags (8 cm x 8 cm in size; 55 µm nylon screen), which allowed entry of fungal hyphae, bacteria, and soil microfauna while minimizing particle loss 2 . The mesh bags were sealed with hot glue and shipped back to the sites of origin and buried at a depth of 0–15 cm at the same locations where soils were initially sampled, and geo-referenced to facilitate retrieval. The mesh bags with litter + unlabeled lignin were buried at even-numbered sampling points for each site, and those with litter + 13 C β -labeled lignin were buried at the odd-numbered sampling points. After approximately 1 y of field incubation, the mesh bags were retrieved by NEON staff, flash-frozen on dry ice, and shipped on ice to ISU. Some bags were damaged or could not be located in the field (31 out of 320 samples). The soil and litter mixture was subsampled from each mesh bag, and then air-dried and finely ground for analysis of C concentrations and δ 13 C at the UC Davis Stable Isotope Facility using an elemental analyzer (Elementar Analysensysteme GmbH, Hanau, Germany) and continuous flow isotope ratio mass spectrometer (Sercon Ltd., Cheshire, UK). Lignin C remaining after the 1-y field incubation was calculated by multiplying f lignin calculated based on two-source mixing model (see details in SI) by the total C concentration in samples from the soil + litter + 13 C β -labeled lignin treatment, with corrections accounting for new C inputs as necessary based on measurements of the samples with unlabeled lignin (see details in SI). Soil inorganic N availability. We measured ammonium (NH 4 + ) and nitrate (NO 3 − ) in additional replicate soil + litter mixture samples (10:1 mass ratio of soil to litter) from all soils used in the lab incubation after 1, 9, and 18 months. Briefly, 10 g soil mixed with 1 g litter was placed in a 50 mL centrifuge tube, loosely covered, and then incubated at 23°C in the dark after adjusting soil moisture to field capacity. Water was periodically added to soil samples to replace vapor loss, measured gravimetrically. Soil (~ 2g) was subsampled from each centrifuge tube and extracted with 2 M potassium chloride at each timepoint. The soil solution was analyzed by microplate colorimetry for NH 4 + -N 60 . The NO 3 − -N was analyzed by microplate colorimetry 61 or for the 9-month samples, second-derivative spectroscopy 62 ; these methods agreed almost perfectly on a subset of samples (slope = 0.95, R 2 = 0.97). Net N mineralization was calculated as the difference in inorganic N between sets of sampling points (9-month vs. 1-month; 18-month vs. 9-month; 18-month vs. 1-month). Soil geochemical analysis. Most physical and geochemical measurements were made on soils from all of the sampling points used for the field and lab incubations, except for particle size and 0.5 M HCl extractions, which were done for the four sampling points per site used for laboratory incubation. Physical and geochemical measurements included soil pH, particle size fractions, 0.5 M HCl-extractable Fe(II) and Fe(III), ammonium oxalate-extracted metals (Al, Fe, Mn), and citrate dithionite-extracted metals (Al, Fe, Mn and Ca). Some of these data were presented previously in a manuscript describing relationships between soil properties and particulate and mineral-associated organic matter fractions of these soils 53 . Field-moist soil subsamples were measured for pH in 1:1 slurries of soil and deionized water. Air-dried subsamples were used to measure particle size (sand, silt and clay) by sieving and sedimentation following aggregate dispersion with sodium hexametaphosphate 53 . Field-moist subsamples were extracted with 0.5 M hydrochloric acid (HCl) to measure dissolved and adsorbed Fe(II) as well as dissolved, organically-complexed Fe(III), and a highly reactive fraction of Fe(III) minerals 63 . Concentrations of Fe(II) and Fe(III) were measured colorimetrically 64 and summed as Fe HCl . Additional air-dried subsamples were extracted with acid ammonium oxalate in the dark at pH 3 to measure organo-metal complexes and short-range-ordered (SRO) phases of Al, Fe, and Mn (denoted Al ox , Fe ox , Mn ox ), and with sodium citrate dithionite to measure the crystalline and SRO phases of Fe (Fe cd ) as well co-occurring Al, Mn, and Ca (Al cd , Mn cd , and Ca cd ) 65 . Metals were analyzed via inductively coupled plasma optical emission spectrometry (PerkinElmer Optima 5300 DV, Waltham, MA). Extraction of Al and Mn by oxalate and citrate-dithionite were very similar, so we only report Al ox and Mn cd . The difference between Fe cd and Fe ox represents crystalline phases (Fe cd−ox ). We interpret Mn cd as including exchangeable Mn, organo-metal complexes, and poorly crystalline phases. We interpret Ca cd as a measure of exchangeable Ca and Ca in organo-Fe associations 66 . Microbial analysis. DNA was extracted from soils for ITS amplicon sequencing and quantitative PCR of ITS and 16S regions. Each of the four soils per site used for lab incubation was subsampled for DNA extraction at the beginning of the incubation, and additional replicates were extracted after 9 and 14 months. The incubated replicates used for DNA extraction were prepared similarly to the replicates used for CO 2 analyses, and were amended with A. gerardi litter in a 1:10 mass ratio of litter to soil. The field-incubated soils corresponding to the same four sampling points for each site used in the lab incubation were also extracted for DNA, totaling 548 samples overall (156 soils × 3 time points for lab incubation and 80 soils for field incubation). Soils were stored at -80°C before DNA extraction from 250 mg subsamples using the MagAttract PowerSoil DNA EP Kit (Qiagen, USA) on an Eppendorf epMotion 5075 liquid handling robot (Eppendorf North America, USA). Concentrations of DNA were measured using a Quant-iT™ dsDNA high sensitivity Assay Kit (Invitrogen, USA) to standardize DNA masses for sequencing. Samples were diluted to 10 ng DNA µL − 1 prior to sequencing; samples with concentration < 10 ng DNA µL − 1 were submitted directly. The ITS1 region of the ITS rRNA gene was amplified using the primer sets ITS1f (CTTGGTCATTTAGAGGAAGTAA) and ITS2 (GCTGCGTTCTTCATCGATGC), with PCR conditions as follows: 1 min at 94°C, followed by 35 cycles of 30 s at 94°C, 30 s at 52°C and 30 s at 68°C, and 10 min at 68°C. Fungal ITS rRNA gene amplicon sequencing was performed on the Illumina Miseq platform at Argonne National Laboratory with library preparation using the Miseq Reagent Kit V2 (Illumina, USA), producing 2 × 250-bp reads. The DNA sequencing data are available at National Center for Biotechnology Information (NCBI) Sequence Read Archive PRJNA808104. Quantitative real-time PCR was performed on a CFX96™ real-time system coupled to a C1000™ thermal cycler (Bio-Rad, USA) to assess the quantity of 16S rRNA and ITS genes. Each sample was prepared using 10 µL of SsoFast EvaGreen Supermix, 0.6 µL of each primer, 2 µL of diluted DNA sample, and nuclease-free water to a final volume of 20 µL. Bacterial 16S rRNA genes were amplified using the primer sets 1055YF(ATGGYTGTCGTCAGCT) and 1392R (ACGGGCGGTGTGTAC) and the following PCR conditions: 2 min at 50°C and 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 58°C 67 . Fungal ITS rRNA genes were amplified using the primer sets ITS1F_KYO1 (CTHGGTCATTTAGAGGAASTAA) and ITS2_KYO2 (TTYRCTRCGTTCTTCATC) 68 and the following PCR conditions: 2 min at 50°C and 2 min at 95°C, followed by 40 cycles of 30 s at 95°C, 30 s at 55°C and 60 sec at 72°C, and 10 min at 72°C. Standard curves for 16S rRNA and ITS rRNA genes were constructed using serial 10-fold dilutions from 10 − 1 to 10 − 8 of known concentrations of synthesized oligonucleotides (Integrated DNA Technologies, USA). Bioinformatics. We used the Divisive Amplicon Denoising Algorithm 2 (DADA2) pipeline 69 to process the ITS sequencing data in R statistical software version 3.6.1 70 . We excluded samples with small numbers of reads (≤ 900 sequences), including 27, 78, and 1 samples collected after 0-, 9-, and 14-months of the lab incubation, respectively. All functions were run using default parameters suggested by the DADA2 pipeline tutorial. The end product included an amplicon sequence variant (ASV) table recording the number of times each exact ASV was observed in each sample, along with a taxa table recording taxonomy assigned to the ASVs from kingdom to species levels, using the naive Bayesian classifier algorithm and the UNITE ITS database version 10.05.2021. Most ASVs had 251–336 bp, falling within the commonly amplified ITS1 length of 200–600 bp. Next, we trimmed the ASV tables using the “phyloseq” package (McMurdie and Holmes, 2013) in R. ASVs with < 10 sequences, i.e., rare ASVs, across all samples were removed. Before trimming, there were 22154 total ASVs and 3118076 total sequences across 442 samples; afterwards, there were 15583 total ASVs and 3085446 total sequences. After removing rare ASVs, there were 4 to 126 ASVs (mean = 55) and 441 to 17234 sequences per sample (mean = 6981). Rarefaction curves suggested that sequencing depths were adequate for all samples (data not shown). Statistical Analysis. For the lab incubation, we explored temporal trends in instantaneous C decomposition rate from each C source at each site and in lignin C decomposition rate for each individual sampling point (Supplementary Fig. S2), using generalized additive mixed models (GAMMs), including an autoregressive error term to account for temporal autocorrelation, using the “mgcv” package 71 version 1.8.28 in R 3.6.1. Pairwise correlations between cumulative C decomposition over 6, 12, and 18 months (lignin, litter, soil and field lignin decomposition) and biogeochemical predictors were tested by Pearson correlation. The biogeochemical predictors included several categories, which we define as follows (1) climatic: MAT and MAP; (2) N-related: bulk N, bulk C/N, NH 4 + -N and NO 3 − -N after 1-, 9- and 18-month incubations; (3) geochemical: soil pH, silt + clay, Al ox , Fe ox , Fe cd−ox , Fe HCl , Mn cd , Ca cd ; (4) microbial: fungal composition, fungal Chao1 richness, fungal quantity, bacterial quantity, and fungal-to-bacterial ratio (Supplementary Table S2). In the microbial predictors, fungal composition was represented by the first (PC1) or second (PC2) axis of a principal coordinate analysis of ITS sequencing data on soils subsampled from the lab incubation at 14 months, conducted in the “vegan” package. The species-level abundance table (rather than the ASV table) was used to calculate Hellinger distances among samples before the analysis to alleviate the issue of a sparse matrix with many zero values 72 . The PC2 of fungal species composition was significantly ( P < 0.01) correlated with cumulative lignin and litter decomposition in the lab incubation and thus used as fungal composition predictor. Similarly, the PC1 of fungal species composition was significantly ( P < 0.01) correlated with cumulative SOC decomposition. Overall fungal composition changed little with time during the lab incubation (Supplementary Fig. S5b). Therefore, for subsequent statistical analyses we used the ITS data from samples collected after 14 months of incubation, because only one sample from this time point was excluded from analyses because of low read counts. Fungal richness was represented by the residual of ASV Chao1 index regressed on the square root of the number of total sequences within a sample, a method that accounts for differences in sequencing depth among samples 73 . We used copy numbers of ITS and 16S rRNA genes in the initial soil samples (1 g dry mass equivalent) as indices of fungal and bacterial quantity in our statistical models. Although fungal and bacterial quantities changed throughout the incubation (Supplementary Fig. S5a), including data from 9 and 14 months did not improve model performance. Fungal-to-bacterial ratio was calculated as fungal quantity divided by bacterial quantity. We also investigated whether putative lignin-degrading fungal organisms, i.e., white-, brown-, and soft-rot fungi identified in the FUNGuild database 74 , were linked to lignin decomposition. Similarly, we used PC1 or PC2 of a principal coordinate analysis based on Hellinger distances among samples from the lab incubation at 14 months calculated using only the ASV abundances of the identified “rot” fungi to represent their overall composition. We summed relative abundance of the identified “rot” fungi for each sample. Pearson correlations were then performed among lab lignin decomposition and overall composition and summed abundance of the “rot” fungi. We also used Pearson correlations to examine relationships among lab lignin decomposition and individual fungal genera occurring in more than 10 samples. We further used linear mixed models (LMMs) and random forest models (RFMs) to identify important predictors for cumulative C decomposition (lignin, litter, soil, and field lignin) variables. We included the above-mentioned climatic, N-related, geochemical, and microbial predictors in models of the laboratory incubation decomposition data. Inorganic N predictors from three timepoints could explain some variation in lab litter decomposition in the RFM but including these predictors did not improve model performance or change variable importance of other key predictors. Thus, inorganic N predictors were not retained in the final models, and we conducted the above-mentioned N addition experiment to specifically test the effects of inorganic N on lignin and litter decomposition. For statistical models of field lignin decomposition, we selected the climatic and geochemical predictors described above. We first fit the models including all categories of predictors and found that microbial predictors, silt + clay, and Fe HCl were not important predictors of field lignin decomposition. Therefore, we re-fit the models excluding these candidate predictors because these data were collected only for the field samples from the locations corresponding to the lab incubation. Inorganic N variables in soil + litter mixtures were not measured for field lignin decomposition. In the LMM, homoscedasticity and normality assumptions were met by raw data, except for lab lignin decomposition, which was log10 transformed. To estimate predictor importance, all variables were standardized to a mean of zero and a standard deviation of one to account for magnitude difference. All predictor variables were used as fixed effects and site was included as a random intercept to account for possible intra-site dependence in the LMMs. Adding sampling location as an additional random effect to account for correlations between 0–15 and 15–30 cm samples did not improve model performance. Some candidate predictors were excluded from initial models because of weak pairwise correlations with response variables (usually r 0.50). Some predictors were further removed from final models through comparison of Akaike Information Criterion (AIC) values of nested models using stepwise backward selection. All predictors in the final models exhibited variance inflation factor values < 3 and correlation coefficients -0.70, implying that collinearity was acceptable. The relative contributions of fixed effects were determined by standardized regression coefficient estimates, and their significance was tested by the Wald chi-square test. LMM performance was evaluated by R 2 representing variance explained by only the fixed effects and by the model, respectively. The LMM analyses were conducted with the “lme4” package 75 . We used random forest models (RFMs) to explore possible nonlinear relationships among predictors and C decomposition variables (Breiman 2001). Variables were not standardized for an easier interpretation of the RFM partial dependence plot, which showed the marginal effect of each predictor on the predicted response variable. Lab lignin decomposition was log10 transformed and the same predictors as in the initial LMMs were included in the initial RFMs. Unimportant predictors were removed from models with Z-score < 5 in the “Boruta” package. Some predictors were further removed from final models based on root mean square error (RMSE) and R 2 using repeated (N = 100) cross validation. The RFMs were not overfit as indicated by an overfitting ratio > 10 in the “rfUtilities” package. RFM was applied with 1,000 trees, with other options sticking to default parameters in the “randomForest” package 77 . The RFM performance was evaluated by R 2 , and model overfitting was examined using the “rfUtilities” package. Variable importance was assessed using increase of mean squared error (%IncMSE) when a given variable is randomly permuted; a larger increase in MSE illustrates greater importance of the permuted variable. All statistical analyses and plotting were performed in R statistical software version 3.6.1 70 . Data availability: The data from this study will be available from the Environmental Data Initiative Data Portal upon acceptance. Declarations Acknowledgments We thank all of the NEON staff who contributed to field sampling, and A. Mirabito, S. Tsui, L. James, A. Boyer and H. Craven for lab assistance. This work was funded in part by National Science Foundation grant 1802745 (SJH, SRW, AH, CL) and Office of Biological and Environmental Research of the U.S Department of Energy Great Lakes Bioenergy Research Center grants DE-FC02-07ER64494, DE-SC0012742. Author contributions: S.J.H. and S.R.W. conceived and designed this study. K.E.H. and V.I.T. conducted lignin syntheses. W.H. and W.Y. performed research. W.Y., E.R. and J.Y. conduct microbial analysis. W.Y. and W.H. analyzed the data. W.H., W.Y. and S.J.H. wrote the manuscript. S.R.W., B.Y., K.E.H., C.L. and A.C.H. provided suggestions for substantial revisions. Additional information Supplementary information 1: Preparation of litter and lignin mixture Supplementary information 2: Partitioning of decomposed C sources in the lab incubation Supplementary information 3: Lignin C decomposition in the field incubation Supplementary Figs. S1 to S11 Supplementary Tables S1 to S3 References Competing interests: Authors declare that they have no competing interests. References Melillo, J. M., Aber, J. D. & Muratore, J. F. Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology 63 , 621–626 (1982). Adair, E. C. et al. Simple three-pool model accurately describes patterns of long-term litter decomposition in diverse climates. Glob. Chang. Biol. 14 , 2636–2660 (2008). Bollag, J.-M., Dec, J. & Huang, P. M. Formation mechanisms of complex organic structures in soil habitats. in Advances in Agronomy (ed. Sparks, D. L.) vol. 63 237–266 (Academic Press, MA, 1997). Chen, J. et al. A keystone microbial enzyme for nitrogen control of soil carbon storage. Sci. Adv. 4 , eaaq1689 (2018). Hall, S. J., Huang, W., Timokhin, V. I. & Hammel, K. E. Lignin lags, leads, or limits the decomposition of litter and soil organic carbon. Ecology 101 , e03113 (2020). Talbot, J. M. & Treseder, K. K. Interactions among lignin, cellulose, and nitrogen drive litter chemistry–decay relationships. Ecology 93 , 345–354 (2012). Thevenot, M., Dignac, M.-F. & Rumpel, C. Fate of lignins in soils: A review. Soil Biol. Biochem. 42 , 1200–1211 (2010). Cotrufo, M. F., Wallenstein, M. D., Boot, C. M., Denef, K. & Paul, E. The Microbial Efficiency-Matrix Stabilization (MEMS) framework integrates plant litter decomposition with soil organic matter stabilization: do labile plant inputs form stable soil organic matter? Glob. Chang. Biol. 19 , 988–995 (2013). Hall, S. J., Ye, C., Weintraub, S. R. & Hockaday, W. C. Molecular trade-offs in soil organic carbon composition at continental scale. Nat. Geosci. 13 , 687–692 (2020). Rasmussen, C., Matsuyama, N., Dahlgren, R. A., Southard, R. J. & Brauer, N. Soil genesis and mineral transformation across an environmental gradient on andesitic lahar. Soil Sci. Soc. Am. J. 71 , 225–237 (2007). Doetterl, S. et al. Soil carbon storage controlled by interactions between geochemistry and climate. Nat. Geosci. 8 , 780–783 (2015). Moorhead, D. L. & Sinsabaugh, R. L. A theoretical model of litter decay and microbial interaction. Ecol. Monogr. 76 , 151–174 (2006). Kleber, M. et al. Chapter One - Mineral–organic associations: Formation, properties, and relevance in soil environments. in Advances in Agronomy (ed. Sparks, D. L.) vol. 130 1–140 (Academic Press, MA, 2015). Kaiser, K., Guggenberger, G. & Haumaier, L. Changes in dissolved lignin-derived phenols, neutral sugars, uronic acids, and amino sugars with depth in forested Haplic Arenosols and Rendzic Leptosols. Biogeochemistry 70 , 135–151 (2004). Coward, E. K., Ohno, T. & Plante, A. F. Adsorption and molecular fractionation of dissolved organic matter on iron-bearing mineral matrices of varying crystallinity. Environ. Sci. Technol. 52 , 1036–1044 (2018). Huang, W. et al. Enrichment of lignin-derived carbon in mineral-associated soil organic matter. Environ. Sci. Technol. 53 , 7522–7531 (2019). Li, H., Santos, F., Butler, K. & Herndon, E. A critical review on the multiple roles of manganese in stabilizing and destabilizing soil organic matter. Environ. Sci. Technol. 55 , 12136–12152 (2021). Keiluweit, M. et al. Long-term litter decomposition controlled by manganese redox cycling. PNAS 112 , E5253–E5260 (2015). Rowley, M. C., Grand, S. & Verrecchia, É. P. Calcium-mediated stabilisation of soil organic carbon. Biogeochemistry 137 , 27–49 (2018). Lovett, G. M., Arthur, M. A. & Crowley, K. F. Effects of calcium on the rate and extent of litter decomposition in a northern hardwood forest. Ecosystems 19 , 87–97 (2016). Louca, S. et al. Function and functional redundancy in microbial systems. Nat. Ecol. Evol. 2 , 936–943 (2018). Bugg, T. D., Ahmad, M., Hardiman, E. M. & Singh, R. The emerging role for bacteria in lignin degradation and bio-product formation. Curr. Opin. Biotechnol. 22 , 394–400 (2011). Hammel, K. E. Fungal degradation of lignin. in Driven by Nature: Plant Litter Quality and Decomposition (eds. Cadisch, G. & Giller, K. E.) 33–46 (CAB International, Wallingford, 1997). Janusz, G. et al. Lignin degradation: microorganisms, enzymes involved, genomes analysis and evolution. FEMS Microbiol. Rev. 41 , 941–962 (2017). Schimel, J. & Schaeffer, S. M. Microbial control over carbon cycling in soil. Front. Microbiol. 3 , Article 348 (2012). Osono, T. Decomposing ability of diverse litter-decomposer macrofungi in subtropical, temperate, and subalpine forests. J. For. Res. 20 , 272–280 (2015). Berg, B. & Staaf, H. Decomposition rate and chemical changes of scots pine needle litter. II. Influence of chemical composition. Ecol. Bull. 32 , 373–390 (1980). Kallenbach, C. M., Frey, S. D. & Grandy, A. S. Direct evidence for microbial-derived soil organic matter formation and its ecophysiological controls. Nat. Commun. 7 , 13630 (2016). Sluiter, J. B., Ruiz, R. O., Scarlata, C. J., Sluiter, A. D. & Templeton, D. W. Compositional analysis of lignocellulosic feedstocks. 1. Review and description of methods. J. Agric. Food Chem. 58 , 9043–9053 (2010). Hernes, P. J., Kaiser, K., Dyda, R. Y. & Cerli, C. Molecular trickery in soil organic matter: Hidden lignin. Environ. Sci. Technol. 47 , 9077–9085 (2013). Haider, K. & Martin, J. P. Decomposition in soil of specifically 14 C-labeled model and cornstalk lignins and coniferyl alcohol over two years as influenced by drying, rewetting, and additions of an available C substrate. Soil Biol. Biochem. 13 , 447–450 (1981). Hall, S. J. & Thompson, A. What do relationships between extractable metals and soil organic carbon concentrations mean? Soil Sci. Soc. Am. J. 86 , 195–208 (2022). Yu, W., Weintraub, S. R. & Hall, S. J. Climatic and geochemical controls on soil carbon at the continental scale: interactions and thresholds. Global Biogeochem. Cycles. 35 , e2020GB006781 (2021). Chen, C., Hall, S. J., Coward, E. & Thompson, A. Iron-mediated organic matter decomposition in humid soils can counteract protection. Nat. Commun. 11 , 2255 (2020). Hofrichter, M. Review: lignin conversion by manganese peroxidase (MnP). Enzyme Microb. Technol. 30 , 454–466 (2002). Grossman, J. J., Cavender-Bares, J. & Hobbie, S. E. Functional diversity of leaf litter mixtures slows decomposition of labile but not recalcitrant carbon over two years. Ecol. Monogr. 90 , e01407 (2020). Berg, B. et al. Maximum decomposition limits of forest litter types: a synthesis. Can. J. Bot. 74 , 659–672 (1996). Kleber, M. et al. Dynamic interactions at the mineral–organic matter interface. Nat. Rev. Earth Environ. 2 , 402–421 (2021). Frankland, J. C. Fungal decomposition of bracken petioles. J. Ecol. 57 , 25–36 (1969). Zare-Maivan, H. & Shearer, C. A. Extracellular enzyme production and cell wall degradation by freshwater lignicolous fungi. Mycologia 80 , 365–375 (1988). Fukasawa, Y. & Matsukura, K. Decay stages of wood and associated fungal communities characterise diversity–decomposition relationships. Sci. Rep. 11 , 8972 (2021). Talbot, J. M. et al. Endemism and functional convergence across the North American soil mycobiome. PNAS 111 , 6341–6346 (2014). Boer, W. de, Folman, L. B., Summerbell, R. C. & Boddy, L. Living in a fungal world: impact of fungi on soil bacterial niche development. FEMS Microbiol. Rev. 29 , 795–811 (2005). Schimel, J. P. Ecosystem consequences of microbial diversity and community structure. in Arctic and alpine biodiversity (eds. Chapin III, F. S. & Korner, C.) 239–244 (Springer-Verlag, Berlin, 1995). Cleveland, C. C. et al. Litter quality versus soil microbial community controls over decomposition: a quantitative analysis. Oecologia 174 , 283–294 (2014). Strickland, M. S., Lauber, C., Fierer, N. & Bradford, M. A. Testing the functional significance of microbial community composition. Ecology 90 , 441–451 (2009). Liu, Y.-R. et al. New insights into the role of microbial community composition in driving soil respiration rates. Soil Biol. Biochem. 118 , 35–41 (2018). Kemmitt, S. J. et al. Mineralization of native soil organic matter is not regulated by the size, activity or composition of the soil microbial biomass—a new perspective. Soil Biol. Biochem. 40 , 61–73 (2008). Waring, B. G. et al. From pools to flow: The PROMISE framework for new insights on soil carbon cycling in a changing world. Glob. Chang. Biol. 26 , 6631–6643 (2020). Strickland, M. S., Keiser, A. D. & Bradford, M. A. Climate history shapes contemporary leaf litter decomposition. Biogeochemistry 122 , 165–174 (2015). Glassman, S. I. et al. Decomposition responses to climate depend on microbial community composition. PNAS 115 , 11994–11999 (2018). Bradford, M. Thermal adaptation of decomposer communities in warming soils. Front. Microbiol. 4 , Article 333 (2013). Yu, W., Huang, W., Weintraub-Leff, S. R. & Hall, S. J. Where and why do particulate organic matter (POM) and mineral-associated organic matter (MAOM) differ among diverse soils? Soil Biol. Biochem. 172 , 108756 (2022). Rasmussen, C. et al. Beyond clay: towards an improved set of variables for predicting soil organic matter content. Biogeochemistry 137 , 297–306 (2018). Wagai, R., Kajiura, M. & Asano, M. Iron and aluminum association with microbially processed organic matter via meso-density aggregate formation across soils: organo-metallic glue hypothesis. SOIL 6 , 597–627 (2020). Keller, M., Schimel, D. S., Hargrove, W. W. & Hoffman, F. M. A continental strategy for the National Ecological Observatory Network. Front. Ecol. Environ. 6 , 282–284 (2008). SanClements, M. et al. Collaborating with NEON. BioScience 70 , 107–107 (2020). Hall, S. J., Huang, W. & Hammel, K. E. An optical method for carbon dioxide isotopes and mole fractions in small gas samples: Tracing microbial respiration from soil, litter, and lignin. Rapid Commun. Mass Spectrom. 31 , 1938–1946 (2017). Lawrence, N. C., Tenesaca, C. G., VanLoocke, A. & Hall, S. J. Nitrous oxide emissions from agricultural soils challenge climate sustainability in the US Corn Belt. PANS 118 , e2112108118 (2021). Weatherburn, M. W. Phenol-hypochlorite reaction for determination of ammonia. Anal. Chem. 39 , 971–974 (1967). Doane, T. A. & Horwáth, W. R. Spectrophotometric determination of nitrate with a single reagent. Anal. Lett. 36 , 2713–2722 (2003). Crumpton, W. G., Isenhart, T. M. & Mitchell, P. D. Nitrate and organic N analyses with second-derivative spectroscopy. Limnol. Oceanogr. 37 , 907–913 (1992). Hall, S. J. & Silver, W. L. Reducing conditions, reactive metals, and their interactions can explain spatial patterns of surface soil carbon in a humid tropical forest. Biogeochemistry 125 , 149–165 (2015). Huang, W. & Hall, S. J. Optimized high-throughput methods for quantifying iron biogeochemical dynamics in soil. Geoderma 306 , 67–72 (2017). Loeppert, R. H. & Inskeep, W. P. Iron. in Methods of Soil Analysis 639–664 (John Wiley & Sons, Ltd, Hoboken, NJ, 1996). Hall, S. J. & Huang, W. Iron reduction: a mechanism for dynamic cycling of occluded cations in tropical forest soils? Biogeochemistry 136 , 91–102 (2017). Ritalahti, K. M. et al. Quantitative PCR targeting 16S rRNA and reductive dehalogenase genes simultaneously monitors multiple Dehalococcoides strains. Appl. Environ. Microbiol. 72 , 2765–2774 (2006). Toju, H., Tanabe, A. S., Yamamoto, S. & Sato, H. High-coverage ITS primers for the DNA-based identification of Ascomycetes and Basidiomycetes in environmental samples. PLoS One 7 , e40863 (2012). Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods 13 , 581–583 (2016). R Core Team. R: A language and environment for statistical computing . (R Foundation for Statistical Computing, Vienna, 2019). Wood, S. N. Generalized Additive Models: An Introduction with R, Second Edition . (Chapman and Hall/CRC, NY, 2017). Legendre, P. & Gallagher, E. D. Ecologically meaningful transformations for ordination of species data. Oecologia 129 , 271–280 (2001). Tedersoo, L. et al. Global diversity and geography of soil fungi. Science 346 , 1256688 (2014). Nguyen, N. H. et al. FUNGuild: An open annotation tool for parsing fungal community datasets by ecological guild. Fungal Ecol. 20 , 241–248 (2016). Bates, D., Maechler, M., Bolker, B. & Walker, S. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67 , 1–48 (2015). Breiman, L. Random forests. Machine Learning 45 , 5–32 (2001). Liaw, A. & Wiener, M. Classification and regression by random forest. R News 2 , 18–22 (2002). Additional Declarations There is NO Competing Interest. Supplementary Files ligninSINCWenjuans091622.docx Supplementary Information Cite Share Download PDF Status: Published Journal Publication published 19 Apr, 2023 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-2086399","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":139078870,"identity":"68348dc5-5870-42c7-87ce-19ad8c50f4da","order_by":0,"name":"Wenjuan Huang","email":"","orcid":"https://orcid.org/0000-0003-1038-1591","institution":"Iowa State University","correspondingAuthor":false,"prefix":"","firstName":"Wenjuan","middleName":"","lastName":"Huang","suffix":""},{"id":139078871,"identity":"7bd9c543-3be7-4f0e-a6a2-e7da378255e6","order_by":1,"name":"Wenjuan Yu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAv0lEQVRIiWNgGAWjYFCCAwwMHwzgPGbitDDOAGrhgagmSgtQGQ8DKVoMDp4xk7YpsLO3Zz9/8AFDhXViA0EtB4BacgySE3t4kpkNGM6kE9ZiBtHCnMDDkMwmwdh2mEgtFgb19jz8j9l/MP4jVguDwWHGHolkNgbGBiK02B84VmzZY3A8sefGY2OJhGPpxgS1SM44vPHGjz/V9uz9iQ8/fKixliWohUHiBCLqGRIIKgcB/vYHRKkbBaNgFIyCEQwAQhY8cCjsM6EAAAAASUVORK5CYII=","orcid":"","institution":"Iowa State University","correspondingAuthor":true,"prefix":"","firstName":"Wenjuan","middleName":"","lastName":"Yu","suffix":""},{"id":139078872,"identity":"af8c9f2f-8859-417a-aed7-ab54b8323d7a","order_by":2,"name":"Bo Yi","email":"","orcid":"https://orcid.org/0000-0002-8674-4400","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Bo","middleName":"","lastName":"Yi","suffix":""},{"id":139078873,"identity":"a1e1b581-91ea-4d7f-809a-a4fb37f8b4bd","order_by":3,"name":"Erik Raman","email":"","orcid":"","institution":"Iowa State University","correspondingAuthor":false,"prefix":"","firstName":"Erik","middleName":"","lastName":"Raman","suffix":""},{"id":139078874,"identity":"ab40dffd-6f51-49c7-a870-cb6cc64c48e3","order_by":4,"name":"Jihoon Yang","email":"","orcid":"","institution":"Iowa State University","correspondingAuthor":false,"prefix":"","firstName":"Jihoon","middleName":"","lastName":"Yang","suffix":""},{"id":139078875,"identity":"be03bcf5-9092-4587-b197-6b7aa30ee32f","order_by":5,"name":"KE Hammel","email":"","orcid":"","institution":"Univ Wisconsin","correspondingAuthor":false,"prefix":"","firstName":"KE","middleName":"","lastName":"Hammel","suffix":""},{"id":139078876,"identity":"67e376d4-2680-466e-862e-14228308eff3","order_by":6,"name":"Vitaliy Timokhin","email":"","orcid":"","institution":"University of Wisconsin–Madison","correspondingAuthor":false,"prefix":"","firstName":"Vitaliy","middleName":"","lastName":"Timokhin","suffix":""},{"id":139078877,"identity":"261ca693-8b44-4b81-8645-840806cb2bc3","order_by":7,"name":"Chaoqun Lu","email":"","orcid":"https://orcid.org/0000-0002-1526-0513","institution":"Iowa State University","correspondingAuthor":false,"prefix":"","firstName":"Chaoqun","middleName":"","lastName":"Lu","suffix":""},{"id":139078878,"identity":"262c02fb-ad51-4821-912a-14953ef4f11d","order_by":8,"name":"Adina Howe","email":"","orcid":"","institution":"Iowa State University","correspondingAuthor":false,"prefix":"","firstName":"Adina","middleName":"","lastName":"Howe","suffix":""},{"id":139078879,"identity":"d9531c4e-741a-4f70-bd43-d565e3f16504","order_by":9,"name":"Samantha Weintraub-Leff","email":"","orcid":"https://orcid.org/0000-0003-4789-5086","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Samantha","middleName":"","lastName":"Weintraub-Leff","suffix":""},{"id":139078880,"identity":"c300a6c7-4ac2-4179-8bad-fa597b8d84aa","order_by":10,"name":"Steven Hall","email":"","orcid":"https://orcid.org/0000-0002-7841-2019","institution":"Iowa State University","correspondingAuthor":false,"prefix":"","firstName":"Steven","middleName":"","lastName":"Hall","suffix":""}],"badges":[],"createdAt":"2022-09-20 21:05:41","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-2086399/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-2086399/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-023-37862-6","type":"published","date":"2023-04-19T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":26899811,"identity":"021cb0c1-0739-4d62-b139-ad320fb7ce29","added_by":"auto","created_at":"2022-09-23 19:00:58","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":845207,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSampling sites and experimental design of this study\u003c/strong\u003e. (a) The 20 NEON sampling sites; (b) soil sampling points around a 40 x 40-m plot at each site; (c) representative variation in biogeochemical predictors, which included climate, N-related, geochemical, and microbial variables (Supplementary Table S2); and (d) experimental design to partition sources of C decomposition. For lab incubation, four mineral soils at 0–15 and 15–30 cm (e.g., points in red, panel b) were incubated with three substrate treatments (soil alone, soil + C4 litter + unlabeled lignin, and soil + C4 litter + 13Cβ-labeled lignin) to partition C decomposition among lignin, litter, and SOC using C stable isotope measurements of CO2. For field incubation, mesh bags with litter + unlabeled lignin or litter + 13Cβ-labeled lignin were buried and retrieved after approximately 1 y to quantify cumulative lignin decomposition.\u003c/p\u003e","description":"","filename":"floatimage12.png","url":"https://assets-eu.researchsquare.com/files/rs-2086399/v1/b9ff04d0e2f65d573d092ea4.png"},{"id":26899363,"identity":"d6b3f924-9547-402a-acc9-070c0f66c60a","added_by":"auto","created_at":"2022-09-23 18:55:58","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":213820,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCarbon decomposition from lignin, litter and soil in lab-incubated samples from 20 NEON sites, expressed as a percentage of the initial C mass in each pool.\u003c/strong\u003e Note the base-10 logarithmic y-axis scale for C decomposition rate (a, c, and e). M1 and M2 denote mineral soil samples from 0-15 and 15-30 cm, respectively. Total C decomposition represents cumulative C decomposition over 18 months. Lines are fit by generalized additive mixed models (GAMMs). Each point represents mean decomposition rate from four sampling points at each site.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-2086399/v1/dd68ef7ba6c971c7d38d6003.png"},{"id":26898551,"identity":"df57adcb-b8cb-40f8-be6a-521ca0fce0ad","added_by":"auto","created_at":"2022-09-23 18:50:58","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":170831,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffects of N addition on cumulative decomposition of lignin and litter incubated in the lab over 18 months, averaged from four sampling points at 0–15 cm depth from 20 NEON sites. \u003c/strong\u003eThe decomposition is expressed as a percentage of the initial C mass in each pool. Numbers correspond to means from each NEON site according to the legend, denoted by four-letter site IDs (Fig. 1 and Supplementary Table S1). Numbers in orange and green denote decomposition of lignin and litter, respectively.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-2086399/v1/54dfbf625987e92c55619277.png"},{"id":26898553,"identity":"265e5379-6e42-449e-af05-e46d738a4e45","added_by":"auto","created_at":"2022-09-23 18:50:58","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":30439,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePairwise correlations of cumulative C decomposition from lignin, litter and SOC with climatic, N-related, geochemical and microbial predictors.\u003c/strong\u003e Significant (P \u0026lt; 0.05) variables are shown in circles. Circle size and darkness indicates relationship strength; circle color indicates relationship direction (blue is positive, red is negative). The NA indicates that inorganic N data was not available for the field incubation.\u003c/p\u003e","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-2086399/v1/6325af12eaccda9f2fea5c7a.png"},{"id":26898256,"identity":"7f3c287e-6600-4974-89af-8c97ebabfc2c","added_by":"auto","created_at":"2022-09-23 18:45:58","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":40729,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffects of predictors on cumulative lignin, litter and SOC decomposition estimated from linear-mixed models (LMM, left) and random forest models (RFM, right).\u003c/strong\u003e R2fixed and R2model represent variance explained by fixed effects and fixed + random effects in the LMM, respectively; R2RFrepresents variance explained by the RFM. For lignin decomposition, R2fixed= 0.43, R2model = 0.45, R2RF = 0.51; for litter decomposition, R2fixed = 0.49, R2model= 0.52, R2RF = 0.60; for soil C decomposition, R2fixed= 0.43, R2model = 0.71, R2RF = 0.55; for field lignin decomposition, R2fixed = 0.31, R2model= 0.53, R2RF = 0.39. Error bars in the LMMs represent confidence intervals (± 2 standard errors) of standardized regression coefficients. %IncMSE in the RFMs show the increase of the mean squared error when a given predictor is randomly permuted; the larger the value, the more important the predictor.\u003c/p\u003e","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-2086399/v1/b5b73f8c35cc7288dc32d7c5.png"},{"id":36044671,"identity":"0338a8df-3e70-456c-9961-970b9bb511dd","added_by":"auto","created_at":"2023-04-20 07:10:51","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2233788,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-2086399/v1/f14c1400-2d52-498c-9228-46c0ace61934.pdf"},{"id":26898260,"identity":"a28621c5-952a-4a06-ba98-d7e08b4d4fa6","added_by":"auto","created_at":"2022-09-23 18:45:59","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":9710518,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"ligninSINCWenjuans091622.docx","url":"https://assets-eu.researchsquare.com/files/rs-2086399/v1/e6a85f9687183a51a7d6bacd.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Contrasting geochemical and fungal controls on decomposition of lignin and soil carbon at continental scale","fulltext":[{"header":"Introduction","content":"\u003cp\u003eLignin is one of the most abundant biopolymers in the terrestrial biosphere and protects other components of plant tissue from microbial attack. Lignin decomposition is critically linked to our understanding of the fate of plant litter, and it might also influence the composition and persistence of soil organic carbon (SOC). However, lignin\u0026rsquo;s importance in controlling litter and SOC decomposition remains controversial. It was traditionally thought that lignin limits overall litter decay after unprotected C compounds have been depleted \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e and that lignin derivatives are major constituents of SOC \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. However, recent work indicated that lignin might decompose fastest during early stages of litter decomposition, perhaps as a consequence of co-metabolic degradation with labile C substrates \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, and that availability of labile C may increase overall lignin loss \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Analyses of extractable lignin phenols suggested that lignin might even decompose faster than bulk SOC \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, and a recent framework proposed that lignin-derived C is less persistent in soil than microbial-derived C \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. The contradictory findings related to lignin decomposition might be related to biogeochemical differences among ecosystems. The persistence of lignin relative to other C substrates might vary systematically with geochemical and microbial characteristics across diverse soils \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. This is because the controls on lignin decomposition are not necessarily the same as controls on bulk litter or SOC decomposition, yet these decomposition processes have rarely been investigated together.\u003c/p\u003e \u003cp\u003eClimate and the ratio of lignin to nitrogen (N) can effectively predict litter decomposition at site to continental scales \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, but these variables may have different relationships with lignin or SOC decomposition. For example, although high precipitation generally increases litter decomposition \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, it can decrease decomposition of lignin or SOC by increasing mineral weathering and chemical stabilization of C with reactive metals \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Greater litter N content may increase lignin and litter decomposition by alleviating microbial N limitation \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e, whereas increased N availability may also decrease decomposition of lignin or SOC by suppressing the production of oxidative enzymes \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Both mechanisms may occur in the same soils depending on the stage of litter decomposition; N may stimulate early stages while inhibiting later stages of litter decay \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. A consistent measurement of lignin decomposition across greatly differing soils would enhance our understanding of the roles of climate and N availability in regulating decomposition and their relative importance as compared to other soil properties.\u003c/p\u003e \u003cp\u003eIn addition to the traditional variables of climate and N availability, soil geochemical characteristics might also be important predictors of lignin and litter decomposition in ways that differ from bulk SOC. Soil minerals and metals can protect SOC from microbial decomposition through sorption, co-precipitation and polyvalent cation bridging \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. In some soils, lignin-derived C may preferentially associate with Fe and Al relative to bulk litter or bulk SOC \u003csup\u003e\u003cspan additionalcitationids=\"CR15\" citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Correspondingly, reactive forms of Fe and Al might be more important for limiting decomposition of lignin vs. other compounds in litter or bulk SOC. Similar to Fe and Al, Mn might protect C by physical or chemical mechanisms \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, yet because some forms of Mn are powerful oxidants, increased Mn availability can also stimulate lignin decomposition \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Calcium might also have a contrasting relationship between decomposition of different substrates: Ca promotes physicochemical protection of SOC and often negatively correlates with SOC decomposition \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, yet Ca availability may stimulate lignin-degrading fungi and correlate positively with lignin and litter decomposition \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. However, the consistency of relationships between metals and decomposition rates of lignin, litter, and/or SOC across diverse ecosystem types remains unresolved.\u003c/p\u003e \u003cp\u003eBesides geochemistry, the composition and abundance of soil microbes may also impact decomposition. Many microbial taxa may perform similar functions, and it remains elusive whether soil microbial composition explains variation in process rates \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Yet, microbial composition might differentially impact the decomposition of different C forms. On one hand, due to lignin\u0026rsquo;s complex structure of heterogeneous aromatic subunits linked by non-hydrolyzable bonds, only a small subset of microbial taxa, primarily white-, brown-, and soft-rot fungi, and certain bacteria, has been conclusively demonstrated to cleave the lignin macromolecule at the propyl side chain, which is likely the rate-limiting step in lignin decomposition \u003csup\u003e\u003cspan additionalcitationids=\"CR23\" citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. On the other hand, microbial community composition may be less important for decomposition of SOC than for lignin simply because physical restrictions on microbial access to C substrates may predominantly limit SOC mineralization \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. However, few studies have directly compared the relationships of fungi with decomposition of lignin and litter vs. SOC, and those that did mainly quantified rates of lignin decay along with individual fungal taxa at local scales \u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Fungal community analyses paired with measurements of lignin, litter, and SOC decomposition across a set of heterogenous soils could help us to evaluate how composition, diversity, and abundance of soil fungal communities are linked to decomposition among ecosystems.\u003c/p\u003e \u003cp\u003eHere, we measured decomposition of lignin, bulk litter, and SOC via a uniform and quantitative isotopic method from mineral soil samples collected across broad biophysical gradients to test biogeochemical controls on decomposition of these substrates. We used 20 sites from the National Ecological Observatory Network (NEON) that span diverse ecosystems, climatic zones (tundra to tropics), and soil characteristics across North America (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea\u0026ndash;c, Supplementary Table S1 and Fig. S1). Previous examinations of lignin decomposition have often relied on semi-quantitative or indirect methods, such as acid-unhydrolyzable residue to approximate lignin content \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e, oxidation of simple substrates as a measure of potential ligninolytic enzymes \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, or use of lignin monomers rather than polymers in incubation experiments \u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. These methods can substantially underestimate or overestimate the lignin content of litter and soil as well as the activities of ligninolytic enzymes \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. However, C isotope-labeled, high-molecular-weight synthetic lignins provide a tracer that allows unambiguous quantitative measurement of lignin decomposition \u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. We combined isotope-labeled lignin with natural-abundance litter derived from a C\u003csub\u003e4\u003c/sub\u003e grass and added these mixtures to separate soil samples for incubation in the lab, enabling us to quantify lignin, litter and SOC decomposition over time (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed) \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, and to assess their relationships with climatic, N-related, geochemical and microbial factors across soils. Lignin decomposition was also measured in field-incubated samples. We hypothesized that (1) lignin decomposition predictably varies with soil geochemical characteristics (reactive minerals and metals) and fungal communities at the continental scale, in addition to climatic and N-related variables; and (2) the predictors for lignin decomposition are similar with litter decomposition, while these predictors have different relationships with SOC decomposition due to specific interactions of lignin with metals and microbial communities.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e \u003cb\u003eDecomposition rates of lignin, litter, and SOC.\u003c/b\u003e The temporal dynamics of lignin C decomposition were generally more variable than litter C or SOC decomposition, although instantaneous lignin decomposition was about 4-fold lower than litter and soil decomposition, on average, when normalized by C mass in these pools (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Lignin decomposition rate generally decreased over time; however, in some soils it transiently increased over timescales of months or was still increasing at the end of incubation (after 18 months) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea and Supplementary Fig. S2). Litter decomposition rate also increased transiently over time in some sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec), and it was significantly related to instantaneous lignin decomposition rate as indicated by Pearson correlation (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01). The temporal pattern of lignin C decomposition differed from SOC decomposition, which generally showed a declining trend at the end of the incubation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee); instantaneous decomposition of lignin C and SOC were not statistically related (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). At the end of the lab incubation (571 d), cumulative C decomposition relative to initial C was 1.7\u0026ndash;31.4% for lignin, 2.0\u0026ndash;53.0% for litter and 6.3\u0026ndash;91.2% for SOC (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, d, and f). The cumulative decomposition of lignin and SOC was similar between 0\u0026ndash;15 cm and 15\u0026ndash;30 cm soil samples, while cumulative litter decomposition was significantly lower in the deeper soil (34% for 0\u0026ndash;15 cm vs. 23% for 15\u0026ndash;30 cm; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe also conducted a separate N addition experiment to test the effects of N availability on lignin and litter decomposition in the 0\u0026ndash;15 cm soils. Lignin and litter decomposition rates were slightly increased by N addition in most sites throughout the lab incubation (Supplementary Fig. S3). Overall, cumulative decomposition of lignin and litter was slightly but significantly increased by N addition after the 571-d incubation (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01 for both; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). On average, the N addition treatment increased cumulative lignin decomposition by 1.6% and litter decomposition by 6.2% across the 20 sites. Neither lignin nor litter decomposition was significantly depressed by N addition at any site after 18 months (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe further compared lignin decomposition measured in the lab with field incubations of 0\u0026ndash;15 cm soils. Cumulative lignin C decomposition measured in the field-incubated samples after approximately 1 y showed higher variation among samples and overall higher rates than observed in the lab (Supplementary Fig. S4), corresponding with overall higher fungal quantity in the field than in the lab (Supplementary Fig. S1). Total lignin C loss relative to initial lignin C concentrations averaged 9\u0026ndash;63% across the 20 sites in the field, while the site-averaged lignin C loss in the lab ranged from 3\u0026ndash;16% (Supplementary Fig. S4). There was no significant correlation between field and lab lignin C loss.\u003c/p\u003e \u003cp\u003e \u003cb\u003eVariation in biogeochemical predictors among soils.\u003c/b\u003e Along with climatic factors (mean annual temperature, MAT, and mean annual precipitation, MAP), we selected 25 biogeochemical predictors and separated them into three categories, including those related to N availability (11), geochemistry (8), and microbes (6) (Supplementary Table S2). The variation of these biogeochemical predictors was very high among sites and was generally similar between the lab and field incubation datasets (Supplementary Fig. S1). For the N predictors, total soil N was 0.1\u0026ndash;32 mg g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and C/N was 8\u0026ndash;58. Both NH\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e+\u003c/sup\u003e-N (0\u0026ndash;104 \u0026micro;g N g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and NO\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e-N (0\u0026ndash;937 \u0026micro;g N g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) tended to increase with incubation time in the lab (Supplementary Fig. S1a). For the geochemical predictors, soil pH was 4.0\u0026ndash;9.2, and soil particle size (silt\u0026thinsp;+\u0026thinsp;clay, 5\u0026ndash;91%) and metals had at least one order of magnitude difference among sites (0\u0026ndash;30 mg g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e Al\u003csub\u003eox\u003c/sub\u003e; 0\u0026ndash;20 mg g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e Fe\u003csub\u003eHCl\u003c/sub\u003e, 0\u0026ndash;78 mg g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e Fe\u003csub\u003eox\u003c/sub\u003e; 0\u0026ndash;54 mg g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e, 0.0\u0026ndash;3.3 mg g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e Mn\u003csub\u003ecd\u003c/sub\u003e, 0\u0026ndash;55 mg g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e Ca\u003csub\u003ecd\u003c/sub\u003e; Supplementary Fig. S1b). Microbial community variables exhibited as much as four orders of magnitude difference among sites, with fungal quantity of 1.1\u0026times;10\u003csup\u003e5\u003c/sup\u003e\u0026ndash;2.6\u0026times;10\u003csup\u003e9\u003c/sup\u003e gene copies g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, bacterial quantity of 5.2\u0026times;10\u003csup\u003e9\u003c/sup\u003e\u0026ndash;9.8\u0026times;10\u003csup\u003e11\u003c/sup\u003e gene copies g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, fungal-to-bacterial ratio of 5.8\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e\u0026ndash;6.2\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e, and a fungal diversity index of -39\u0026ndash;73 (i.e., the residual of the Chao1 index for ITS ASV\u0026rsquo;s, Supplementary Fig. S1c). Microbial quantity changed with time: fungal quantity increased in the 9-month incubated samples vs. initial soil samples, and fungal quantity further increased and bacterial quantity also increased after 14 months vs. 9 months (Supplementary Fig. S5a).\u003c/p\u003e \u003cp\u003e \u003cb\u003eRelationships of C decomposition with biogeochemical predictors.\u003c/b\u003e Cumulative decomposition of lignin and litter at the end of lab incubation tended to have similar relationships with biogeochemical predictors, whereas lab lignin and SOC decomposition had generally opposite pairwise relationships with those same biogeochemical predictors (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). We assessed these relationships after 6, 12 and 18 months of cumulative decomposition (Supplementary Figs. S6 and S7), and they changed little over time, such that we focused our subsequent analysis on the 18-month (571 d) dataset. Different axes of fungal community composition correlated with lignin and litter vs. SOC decomposition. MAP was significantly and positively related to lignin decomposition but negatively related to SOC decomposition (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01). Lignin decomposition had a significantly negative relationship with NO\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e-N after 1 m of incubation (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01), while litter decomposition was positively related to total N, NH\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e+\u003c/sup\u003e-N, NO\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e-N and inorganic N (NH\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e+\u003c/sup\u003e-N\u0026thinsp;+\u0026thinsp;NO3\u003csup\u003e\u0026minus;\u003c/sup\u003e-N) after 18 months (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 for NH\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e+\u003c/sup\u003e-N and \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01 for the other variables). Soil pH was significantly and negatively related with lignin decomposition but positively related with SOC decomposition (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01). The silt\u0026thinsp;+\u0026thinsp;clay content and some soil metals (Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e and Mn\u003csub\u003ecd\u003c/sub\u003e) had significant and positive relationships with lignin decomposition but negative relationships with SOC decomposition (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01). Fungal quantity had a positive relationship with lignin and litter decomposition but a negative relationship with SOC decomposition. Compared with lab lignin decomposition, field lignin decomposition was also closely related to some predictors within climatic, N, geochemical and microbial categories, although not all relationships were consistent. Specifically, fungal quantity had a positive relationship with field lignin decomposition, consistent with the lab lignin decomposition. However, the field lignin decomposition had a negative relationship with MAP and a positive relationship with soil pH, which was inconsistent with the lab lignin decomposition.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eImportance of biogeochemical predictors for C decomposition.\u003c/b\u003e We used two statistical model approaches (linear mixed model, LMM and random forest model, RFM) to identify the most important soil geochemical, microbial, N, and climatic predictors of decomposition (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The RFM partial dependence plots showed that many relationships between predictors and response variables were approximately linear until predictors increased above the 90th or larger percentiles, where response variables became approximately constant (Supplementary Figs. S8 to S11). The LMM showed that soil pH and fungal composition were the strongest predictors for lab lignin decomposition, and that MAT, fungal quantity, Mn\u003csub\u003ecd\u003c/sub\u003e, Ca\u003csub\u003ecd\u003c/sub\u003e, silt\u0026thinsp;+\u0026thinsp;clay, Fe\u003csub\u003eox\u003c/sub\u003e, and soil C/N could also improve the final model (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Soil pH and Fe\u003csub\u003eox\u003c/sub\u003e had negative relationships with lab lignin decomposition while all other predictors had positive relationships. These predictors explained 43% of the observed variance in lab lignin decomposition; the overall model (including random effects for site) explained 45%. A slightly higher proportion of variation in lab lignin decomposition (R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.51) was explained in the RFM. The optimal RFM included the same predictors as the LMM (except for Fe\u003csub\u003eox\u003c/sub\u003e), with the addition of bacterial quantity, MAP, and Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e. Predicted lignin decomposition decreased with pH and increased with most other predictors (Supplementary Fig. S8). However, predicted lignin decomposition varied nonlinearly with soil C/N, with a minimum at C/N\u0026thinsp;=\u0026thinsp;20. Ten fungal genera occurring in more than 10 samples had significant (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) correlations with lignin decomposition (Supplementary Table S3): positive for \u003cem\u003ePleotrichocladium, Geomyces, Trichocladium, Mycena, Hypochnicium, Solicoccozyma\u003c/em\u003e, and \u003cem\u003eMortierella\u003c/em\u003e and negative for \u003cem\u003eCladophialophora, Pseudofabraea\u003c/em\u003e, and \u003cem\u003eGlarea\u003c/em\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eLignin decomposition in the field and lab shared many of the same predictors in the LMMs, including MAT, Fe\u003csub\u003eox\u003c/sub\u003e, Ca\u003csub\u003ecd\u003c/sub\u003e, and Mn\u003csub\u003ecd\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Fe\u003csub\u003eox\u003c/sub\u003e showed a negative relationship while the other three predictors showed positive relationships with field lignin decomposition. These four predictors, along with MAP, soil C/N, soil N, pH, and Al\u003csub\u003eox\u003c/sub\u003e, explained 31% of the variation in field lignin decomposition; the overall model with random effects explained 53% of the variation. All nine predictors also explained 39% of the variation in field lignin decomposition in the RFM. As observed in the LMM, predicted field lignin decomposition generally increased with increasing Mn\u003csub\u003ecd\u003c/sub\u003e and Ca\u003csub\u003ecd\u003c/sub\u003e and decreased with increasing Fe\u003csub\u003eox\u003c/sub\u003e and Al\u003csub\u003eox\u003c/sub\u003e in the RFM (Supplementary Fig. S9). Unlike lab lignin decomposition, predicted field lignin decomposition showed more complex relationships with MAT with a null relationship below 15\u0026deg;C and a positive relationship at higher temperatures, predicted lignin decomposition decreased with increasing MAP, and Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e was not an important predictor. Similar to the lab lignin decomposition, predicted field lignin decomposition varied nonlinearly with soil C/N, with a minimum at a slightly lower C/N value of 13 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e \u003cp\u003ePredictors of litter decomposition were generally similar to lignin decomposition. The LMM showed that Fe\u003csub\u003eox\u003c/sub\u003e was the strongest predictor of litter decomposition, followed by fungal composition, Mn\u003csub\u003ecd\u003c/sub\u003e, Fe\u003csub\u003eHCl\u003c/sub\u003e, soil N, fungal quantity, bacterial quantity, MAT and pH (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Fe\u003csub\u003eox\u003c/sub\u003e and pH were negatively while all other predictors were positively related to litter decomposition. These predictors explained 49% of the variation in litter decomposition, and the overall model with random effects explained 52%. The optimal RFM explained 60% of the variation in litter decomposition and included the same predictors as the LMM except for silt\u0026thinsp;+\u0026thinsp;clay and soil C/N. Fe\u003csub\u003eox\u003c/sub\u003e, Al\u003csub\u003eox\u003c/sub\u003e, and pH had weak negative relationships with litter decomposition in the RFM while most other predictors had positive relationships (Supplementary Fig. S10).\u003c/p\u003e \u003cp\u003eContrary to lignin and litter decomposition, microbial predictors were not important predictors of SOC decomposition in the statistical models (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The LMM showed that soil C/N, MAT, and silt\u0026thinsp;+\u0026thinsp;clay were the strongest predictors for SOC decomposition, and that MAP, Ca\u003csub\u003ecd\u003c/sub\u003e, Mn\u003csub\u003ecd\u003c/sub\u003e, Al\u003csub\u003eox\u003c/sub\u003e, and pH were also important (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). MAT and pH were positively while all other variables were negatively related to SOC decomposition. The predictors collectively explained 43% of the variation in SOC decomposition and the overall LMM (including random effects) explained 71%. The optimal RFM explained 55% of the variation in SOC decomposition, and included soil C/N and all geochemical and climatic predictors. Predicted SOC decomposition generally decreased with soil C/N, reactive minerals and metals, and MAP, and increased with MAT (Supplementary Fig. S11).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur continental-scale data demonstrated that lignin decomposition was not universally slow or fast when compared with decomposition of litter and SOC, but rather, it varied predictably among sites along with biogeochemical variables (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Consistent with our first hypothesis, a large proportion of variation in lignin decomposition could be predicted by soil geochemical and microbial properties, in addition to the traditional variables of climate and N availability (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Lignin decomposition was highly correlated with litter decomposition and these processes shared the same biogeochemical predictors, but they were generally different from predictors of SOC decomposition (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), supporting our second hypothesis. Several soil geochemical factors had negative (Fe\u003csub\u003eox\u003c/sub\u003e and Al\u003csub\u003eox\u003c/sub\u003e) and positive (Mn\u003csub\u003ecd\u003c/sub\u003e and Ca\u003csub\u003ecd\u003c/sub\u003e) relationships with lignin and/or litter decomposition, while there were not any positive relationships between extractable soil metals and SOC decomposition (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Intriguingly, microbial variables including fungal composition and fungal and bacterial quantity were needed to explain variation in the decomposition of lignin and litter, but not SOC (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). A relatively high amount of N addition stimulated decomposition of lignin and litter to a minor degree in most sites after 18 months (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), but its overall impact was small when considering the wide range of decomposition across samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Inorganic N appeared to be less important than geochemical and microbial properties (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Long-term climatic predictors (MAT and MAP) could explain variation in both field and lab decomposition, but the actual vs. legacy climate, as reflected by the field vs. lab experiments, respectively, had different relationships with decomposition (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOverall, our data showed the particular importance of geochemical and microbial predictors for lignin and litter decomposition, and their differing relationships with SOC decomposition, which collectively supported different aspects of classic and modern views of decomposition. The strong correlation between lignin and litter decomposition and the similar biogeochemical predictors of these processes support the classic view that lignin decomposition is tightly coupled with overall litter decomposition \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. However, we found that decomposition of SOC was unrelated to decomposition of lignin and litter, and that these processes often had contrasting relationships with biogeochemical predictors. This finding is inconsistent with the classic idea that the slow decomposition of lignin residues is an important factor that limits decomposition of total SOC \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. The disparate rates and predictors of lignin and SOC decomposition support the modern notion that lignin depolymerization is not necessarily a primary bottleneck for SOC decomposition \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Furthermore, our dataset provides a mechanistic explanation for the decoupling of lignin and SOC decomposition by highlighting their differing relationships with geochemical and microbial variables.\u003c/p\u003e \u003cp\u003e \u003cb\u003eDiffering geochemical predictors for decomposition of lignin and litter vs. SOC.\u003c/b\u003e We found that geochemical variables often had different relationships with lignin and litter decomposition than with SOC decomposition (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Some geochemical variables had negative (e.g., Fe\u003csub\u003eox\u003c/sub\u003e and Al\u003csub\u003eox\u003c/sub\u003e) and positive (e.g., Fe\u003csub\u003eHCl\u003c/sub\u003e, Mn\u003csub\u003ecd\u003c/sub\u003e and Ca\u003csub\u003ecd\u003c/sub\u003e) relationships with lignin and/or litter decomposition, while others (e.g., Fe\u003csub\u003eHCl\u003c/sub\u003e, Fe\u003csub\u003eox\u003c/sub\u003e, Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e, Al\u003csub\u003eox\u003c/sub\u003e, Mn\u003csub\u003ecd\u003c/sub\u003e, and Ca\u003csub\u003ecd\u003c/sub\u003e) mainly had negative correlations with SOC decomposition. The metals extracted from these NEON soils likely represent ions (e.g. Ca\u003csub\u003ecd\u003c/sub\u003e), metals dissolved from specific mineral phases of varying crystallinity (e.g., Fe\u003csub\u003eHCl\u003c/sub\u003e, Fe\u003csub\u003eox\u003c/sub\u003e, Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e), or a mixture of ions and mineral phases (e.g., Mn\u003csub\u003ecd\u003c/sub\u003e and Al\u003csub\u003eox\u003c/sub\u003e) \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. Soil mineral and metal cations as well as fine particles (silt\u0026thinsp;+\u0026thinsp;clay) are important predictors of SOC concentration due to protection by sorption, precipitation, and aggregation \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. However, such SOC stabilization mechanisms may not necessarily correspond with the short-term decomposition of newly added C sources. Protective effects of soil metals and minerals were mainly applicable for SOC decomposition, and conversely, some of these same variables were actually associated with greater decomposition of lignin and litter.\u003c/p\u003e \u003cp\u003eWe found positive associations of some soil metals (e.g., Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e, Mn\u003csub\u003ecd\u003c/sub\u003e and Ca\u003csub\u003ecd\u003c/sub\u003e) with lignin and litter decomposition (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), consistent with catalytic or biological roles of soil metals for organic matter decomposition demonstrated in other studies \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. The finding that Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e was positively related and Fe\u003csub\u003eox\u003c/sub\u003e was negatively related with lignin and litter decomposition was consistent with multiple functional roles of Fe, which might stimulate decomposition or provide protection depending on C molecular composition and/or redox environment \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. Moreover, in samples with greater Mn and Ca, lignin and litter decomposition increased whereas SOC decomposition decreased (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Our results were consistent with the importance of Mn-promoted degradation and protection of organic C \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, depending on particular C forms. Besides the protective effects of Mn on SOC, Mn also can promote lignin decomposition by stimulating the activities of lignin-degrading enzymes and oxidizing lignin via redox cycling \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, and Mn-stimulated lignin decomposition likely increased overall litter decomposition \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Whereas the negative correlation of SOC decomposition with Ca was likely due to protective cation bridging with soil minerals \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, the strong positive relationships between Ca and decomposition of lignin and litter agreed with previous studies showing that Ca was positively related to the extent of litter mass loss, and in particular, lignin degradation \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Ca is an essential component of the fungal cell wall and can increase the growth of white rot fungi \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. We also found an overall positive relationship of silt\u0026thinsp;+\u0026thinsp;clay with lignin and litter decomposition, which might reflect multiple biological and physical factors that co-vary with particle size, as well as the potential for minerals to catalyze OM decomposition \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Overall, the role of certain metals in stimulating lignin and litter decomposition while suppressing SOC decomposition provides an explanation for fact that these processes may be coupled or decoupled to varying degrees \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, depending on soil characteristics.\u003c/p\u003e \u003cp\u003eIntriguingly, field lignin and lab lignin decomposition had opposite pairwise relationships with several biogeochemical predictors, such that field lignin and lab SOC decomposition also had several biogeochemical predictors in common (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). However, most of these relationships disappeared in the statistical models after other predictors were accounted for (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), and they may have been due to wider ranges of predictor and response variables in the field samples (such as fungal abundance and Fe\u003csub\u003eox\u003c/sub\u003e), and the strikingly different climate conditions in the field than in the lab (Supplementary Figs. S1 and S4).\u003c/p\u003e \u003cp\u003e \u003cb\u003eMicrobial predictors explained decomposition of lignin and litter but not SOC.\u003c/b\u003e Consistent with our hypotheses, composition and quantity of overall fungal communities explained variation in lignin and litter decomposition (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). However, the composition of known \u0026ldquo;rot\u0026rdquo; fungi was insignificantly (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05) correlated with lignin decomposition and summed relative abundance of these fungi had a weak but positive relationship (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.02; data not shown). Seven genera occurring in \u0026gt;\u0026thinsp;10 samples were significantly (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) and positively correlated with lignin decomposition (Supplementary Table S3), but only three have been reported to possibly degrade lignocellulose: \u003cem\u003eTrichocladium\u003c/em\u003e (soft rot), \u003cem\u003eMycena\u003c/em\u003e, and \u003cem\u003eHypochnicium\u003c/em\u003e (white rot) \u003csup\u003e\u003cspan additionalcitationids=\"CR40\" citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. These results indicate that the most commonly studied lignin-degrading fungi (i.e., the known \u0026ldquo;rot\u0026rdquo; fungi) were not necessarily the most important lignin-degrading organisms in our continental-scale dataset \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eConsistent with a previous study across North America \u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, we found that fungal communities were highly heterogenous across sites and even within plots; e.g., only 65 of 342 fungal species occurred in \u0026gt;\u0026thinsp;10 samples. This finding further suggests that specific fungal taxa possibly responsible for lignin decomposition varied with locations and even depths at the same plot. Bacterial quantity was also related to both lignin and litter decomposition (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Although bacteria may degrade lignin directly \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, they might simply be responding to increased C availability as a consequence of fungal lignin decomposition, or may synergistically interact with fungi to promote lignin decomposition \u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. Previous work suggested that microbial community composition does not necessarily influence ecosystem processes because of high functional redundancy \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Yet our findings, in line with previous studies \u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e,\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, clearly showed that differences in fungal community composition and abundance among diverse soils explained variation in lignin and litter decomposition rates when temperature and moisture were held constant. And furthermore, the differing relationships between fungal composition and decomposition of lignin and SOC provide another mechanistic explanation for the observed decoupling of these processes.\u003c/p\u003e \u003cp\u003eContrary to lignin and litter decomposition, microbial variables were less important predictors of SOC decomposition (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), despite high variation in fungal composition and richness and bacterial and fungal quantities across soils (Supplementary Figs. S1 and S5). Different axes of fungal community composition correlated with decomposition of lignin and litter (PC1) vs. SOC (PC2; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The importance of soil microbial communities in driving SOC decomposition remains poorly understood and contrasting findings have been reported. On one hand, significant relationships among SOC decomposition rate and microbial community composition, biomass, and richness have been reported \u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. On the other hand, microbial biomass, community structure or specific activity were weakly related to SOC decomposition \u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. Consistent with the latter findings, we found that despite their significant pairwise correlations (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), microbial predictors including fungal community composition, fungal quantity, and bacterial quantity were not related to SOC decomposition after accounting for other variables (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This finding supports the hypothesized yet not broadly demonstrated view that SOC turnover is dominantly determined by decomposer access to SOC \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Because of the spatially heterogenous soil environment, a large proportion of SOC is probably stored in small pores \u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. The added lignin and litter were likely not protected in those pores, because they were gently mixed into the soil and the particles were much larger than the prevailing pore sizes. This might explain why although C substrates were mixed into soil in this experiment, their decomposition was still measurably related to fungal community composition and quantity.\u003c/p\u003e \u003cp\u003eInterestingly, MAT and MAP of the study sites explained variation in organic matter decomposition (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) even under the common conditions of temperature and moisture imposed in the lab incubation. This is consistent with previous findings that climate history was an important determinant of litter and SOC decomposition, possibly by shaping functional responses of decomposer communities and/or via correlation with soil minerals \u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e,\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. Decomposer communities could adapt to temperature and moisture regimes via changes in enzyme properties and community composition \u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Climate greatly impacts soil weathering \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e, which results in soil geochemical changes that might directly influence OM decomposition, as suggested by our data (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Because the climate variables used as predictors reflected either the actual differences in climate during the field decomposition or the legacies of prior differences in climate during the lab decomposition, respectively, it was perhaps not surprising that MAP had different relationships with lab and field lignin decomposition (Supplementary Figs. S8 and S9). Nevertheless, all organic matter decomposition variables were positively related to MAT in the statistical models (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), suggesting that the legacy effect of soil MAT was stronger than the legacy effect of MAP on OM decomposition.\u003c/p\u003e \u003cp\u003e \u003cb\u003eSmall stimulation of litter and lignin decomposition from increased N availability.\u003c/b\u003e The impact of N on decomposition remains controversial. Whereas greater N availability may increase lignin and litter decomposition rates by alleviating microbial N limitation, it might also decrease late-term decomposition by suppressing oxidative enzymes involved in lignin breakdown \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Despite a negative relationship between initial nitrate and lab lignin decomposition (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), on balance inorganic N addition led to a small net stimulation of lignin and litter decomposition after 18 months (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), and frequently higher decomposition in the N addition vs. control treatments over time (Supplementary Fig. S3). The positive response of lignin and litter decomposition to N addition might imply that microbial growth was N-limited in many sites. Nonlinear relationships between soil C/N and lignin decomposition might reflect multiple mechanisms between N and lignin decomposition (Supplementary Figs. S8 and S9). The positive relationships between total soil N and litter decomposition (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) may be related to greater N availability which could alleviate microbial N limitation and thus facilitate decomposition. The negative relationship between total N and SOC decomposition (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) might result from reduced microbial mining of SOC because of increasing N availability. Overall, despite the stimulation of lignin and litter decomposition, the effects of N on lignin and litter decomposition were relatively small in comparison with variation across sites, even following a substantial addition of inorganic N (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Therefore, when considering continental-scale variation in biogeochemical properties, variation in N availability may be a less important driver of decomposition than sometimes assumed.\u003c/p\u003e \u003cp\u003e \u003cb\u003eImplications for geochemical and microbial control of SOC concentration and composition.\u003c/b\u003e Comparison of our results with other recent observations from NEON soils indicates that the differing controls on decomposition of lignin and litter vs. SOC implied by our data may contribute to variation in SOC concentration and organic matter composition among ecosystems. Many of the same variables that predicted C decomposition in the lab incubation also predicted differences in SOC concentration and the distribution of SOC between SOM size fractions, defined as chemically dispersed particulate organic matter (POM, \u0026gt; 53 \u0026micro;m) and mineral-associated organic matter (MAOM, \u0026lt; 53 \u0026micro;m), which were described in previous studies of NEON soils \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. Silt- and clay-sized minerals and reactive Fe phases in particular have long been thought to protect SOC from decomposition, even though the relationships among these variables can be relatively weak across large datasets \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e. Here, we found that the magnitude or even the \u003cem\u003esign\u003c/em\u003e of the pairwise correlation or model coefficient between decomposition and silt\u0026thinsp;+\u0026thinsp;clay or Fe in various extractions (Fe\u003csub\u003eHCl\u003c/sub\u003e, Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e, Fe\u003csub\u003eox\u003c/sub\u003e) often differed between lignin/litter and SOC (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). These differences could influence SOM composition while explaining the context-dependency of relationships between Fe and SOC concentration in other datasets \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e. For example, negative relationships of Fe\u003csub\u003eox\u003c/sub\u003e with lignin and litter decomposition (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) could help explain the positive relationship between Fe\u003csub\u003eox\u003c/sub\u003e and the increasing proportion of SOC in POM vs. MAOM, which we observed in our previous study with the same soils \u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. That is, Fe\u003csub\u003eox\u003c/sub\u003e could increase POM by disproportionately decreasing rates of lignin decomposition relative to bulk SOC, consistent with the view that POM is mostly composed of decomposing plant detritus which may aggregate with metals \u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e. Similarly, the positive relationship between silt\u0026thinsp;+\u0026thinsp;clay and lignin decomposition and its negative relationship with SOC decomposition is consistent with our previous finding that increased silt\u0026thinsp;+\u0026thinsp;clay was associated with lower SOC in POM vs. MAOM \u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. The decrease in POM vs. MAOM with increasing silt\u0026thinsp;+\u0026thinsp;clay might simply be due to increased capacity for mineral protection, but it might also be linked to increased catalysis of lignin decomposition by metals and/or minerals in these fine particle fractions \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e; Mn and Ca were both associated with greater decomposition of lignin, but not SOC (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Together, the contrasting relationships of silt\u0026thinsp;+\u0026thinsp;clay and Fe\u003csub\u003eox\u003c/sub\u003e with decomposition of lignin and litter vs. SOC provide an explanation for why these variables may be poor predictors of SOC concentration over broad scales, even though they may be related to the physical forms (POM vs. MAOM) of SOC.\u003c/p\u003e \u003cp\u003eIn summary, using a quantitative isotopic method, we found that decomposition of lignin varied 18-fold among soils sampled from sites across North America and incubated in a common environment. Lignin decomposition was always slower than but was strongly related to bulk litter decomposition. Differences in lignin decomposition among sites were strongly related to biogeochemical predictors, in a manner that was similar to bulk litter decomposition but differed from SOC decomposition. Different axes of fungal community composition were related to decomposition of lignin and litter compared to SOC, and metals often positively correlated with lignin decomposition even though they had a neutral or negative correlation with SOC decomposition. Similarities in controls on lignin vs. bulk litter decomposition reinforces the traditional view that lignin is tightly coupled with overall litter decay over timescales of months to years. In contrast, the difference in controls on lignin and litter decomposition vs. SOC supports the modern notion that lignin depolymerization is not a primary bottleneck for SOC decomposition. While substantial research has focused on N dynamics as controls on litter decomposition, our data showed that while significant, the influence of N availability on decomposition of lignin and litter mixed into mineral soils was often smaller than other geochemical and microbial factors. Legacies of previous climates may predict decomposition rates under similar conditions of temperature and moisture. Our data suggest the critical need for mechanistic models to account for contrasting geochemical and microbial controls on decomposition of lignin and litter vs. SOC, in addition to the traditional variables of climate, residue quality, and nutrient availability.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e \u003cb\u003eExperimental design.\u003c/b\u003e We used 20 sites from National Ecological Observatory Network (NEON) to examine decomposition of lignin, bulk litter, and SOC and to test biogeochemical (geochemical, microbial) controls on decomposition of these substrates, in addition to N-related and climatic variables. Soils amended with C stable isotope (\u003csup\u003e13\u003c/sup\u003eC) labeled and un-labeled lignins and a single natural litter source were incubated in the lab to quantify lignin, litter and SOC decomposition over 18 months (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). An additional incubation experiment was also conducted to test the effects of N addition on lignin and litter decomposition. The results of lignin decomposition and its predictors from the lab incubation were further compared with those from a field incubation.\u003c/p\u003e \u003cp\u003e \u003cb\u003eSite selection and soil sampling.\u003c/b\u003e NEON is a U.S. based, continental-scale ecological monitoring network that provides open data, samples, and research infrastructure to reveal how ecosystems are responding to environmental change \u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. For this project, we selected 20 NEON terrestrial sites, denoted by their acronyms as follows: BONA, CPER, DSNY, GRSM, HARV, KONZ, LENO, NIWO, ONAQ, OSBS, PUUM, SJER, SRER, SCBI, TALL, TOOL, UNDE, WREF, WOOD, YELL (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). These sites span wide edaphic, climatic and ecosystem gradients to maximize differences in soil geochemical and microbial communities (Supplementary Fig. S1). They encompass 9 out of the 12 soil orders in the United States Department of Agriculture (USDA) soil classification system (no Histosols, Oxisols, or Vertisols; Supplementary Table S1). The sites had mean annual temperature (MAT) of -9\u0026ndash;22 \u003csup\u003eo\u003c/sup\u003eC and received 262\u0026ndash;2657 mm of mean annual precipitation (MAP) \u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e. The sites included diverse ecosystem types, such as tundra, forest, wetland, grassland, shrubland and desert.\u003c/p\u003e \u003cp\u003eSoils at each site were sampled by NEON staff during the growing season of 2019 (April\u0026ndash;August; later sampling occurred at Alaska sites where soils did not thaw until July or August). Mineral soil samples were collected at two depths (0\u0026ndash;15 cm and 15\u0026ndash;30 cm), after removing any surface litter or organic horizon (Supplementary Table S1), using a 2- to 5-cm diameter corer, according to the standard NEON sampling procedure for that particular site. At each site, samples were collected around the perimeter of one 40 \u0026times; 40-m \u0026ldquo;distributed base plot\u0026rdquo; which was selected to represent the dominant upland vegetation type and soil type of that site, whenever possible, in accordance with site access constraints. Soil at the KONZ site was collected only at 0\u0026ndash;15 cm due to the shallow soil depth. Each plot had 16 replicates (n\u0026thinsp;=\u0026thinsp;16), denoted sampling points 1\u0026ndash;16 hereafter. Point 1 was located 4 m west and 4 m south from the SW corner of each plot, and the other points were located in counterclockwise sequence at 12-m intervals around the perimeter of the plot, each located 4 m outside of the plot boundary (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Soil cores from each point were collected and shipped overnight on ice (~\u0026thinsp;4 \u003csup\u003eo\u003c/sup\u003eC) to Iowa State University (ISU) for use in laboratory and field incubations. Soil from each sample was gently homogenized inside a plastic bag after any coarse roots, macrofauna, or rocks were manually removed. We did not sieve samples except for ONAQ and SRER, where rocks were abundant and were removed by sieving soil through a 2-mm sieve.\u003c/p\u003e \u003cp\u003e \u003cb\u003eLab incubation experiments.\u003c/b\u003e Soils from four sampling points at two depths per site were used for lab incubation and biogeochemical analyses, totaling 156 samples. The four sampling points were mainly selected at the odd number in the middle of each side of the 40 x 40-m plot (red circles in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb), although sampling points from some sites were selected at the numbers next to the middle numbers if soils were not available for both layers. Subsamples of soils used for the lab incubation experiment were brought to field moisture capacity, which was determined for each soil by saturating an additional 20\u0026ndash;30-g subsample placed on filter paper in a funnel, and then measuring gravimetric water content following 48 h of drainage. Subsamples (1 g dry mass equivalent) from each sampling point and depth were incubated under each of three separate substrate treatments to partition C decomposition among three sources, using measurements of d\u003csup\u003e13\u003c/sup\u003eC values of CO\u003csub\u003e2\u003c/sub\u003e. We quantified decomposition of C from extant soil organic matter, C from added litter (senesced leaves of \u003cem\u003eAndropogon gerardi\u003c/em\u003e, a C\u003csub\u003e4\u003c/sub\u003e grass), and a specific C atom (the C\u003csub\u003eβ\u003c/sub\u003e position of the propyl sidechain) in lignin that was precipitated on the added litter. The lignin was prepared as described previously \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. The detailed method is described in the SI. Substrate treatments were: (1) soils alone (control); (2) soils amended with \u003cem\u003eA. gerardi\u003c/em\u003e litter precipitated with trace natural abundance \u003csup\u003e13\u003c/sup\u003eC lignin (soil\u0026thinsp;+\u0026thinsp;litter\u0026thinsp;+\u0026thinsp;unlabeled lignin); (3) soils amended with \u003cem\u003eA. gerardi\u003c/em\u003e litter precipitated with trace lignin labeled with 99 atom percent \u003csup\u003e13\u003c/sup\u003eC at the C\u003csub\u003eβ\u003c/sub\u003e position of each lignin C\u003csub\u003e9\u003c/sub\u003e substructure (soil\u0026thinsp;+\u0026thinsp;litter\u0026thinsp;+\u0026thinsp;\u003csup\u003e13\u003c/sup\u003eC\u003csub\u003eβ\u003c/sub\u003e-labeled lignin). We added uniform litter and synthetic lignin to each of the mineral soil to focus on soil biogeochemical gradients rather than substrate quality. Soils were gently mixed with the litter\u0026thinsp;+\u0026thinsp;lignin mixture in a 250:25:1 ratio of soil:litter:lignin (1 g dry soil mass equivalent was mixed with 100 mg litter and 4 mg lignin). To prepare the litter\u0026thinsp;+\u0026thinsp;lignin mixture, the unlabeled or labeled lignins were precipitated in a 1:25 mass ratio on dried and finely ground leaf litter of \u003cem\u003eA. gerardi\u003c/em\u003e (41.9% C, 0.41% N, and δ\u003csup\u003e13\u003c/sup\u003eC = -12.6\u0026permil;; see SI for more details). The 20 NEON sites comprise ecosystems ranging from C\u003csub\u003e3\u003c/sub\u003e-dominated forest and grassland sites to mixed C\u003csub\u003e3\u003c/sub\u003e\u0026ndash;C\u003csub\u003e4\u003c/sub\u003e grasslands and/or plants with Crassulacean acid metabolism, such that the δ\u003csup\u003e13\u003c/sup\u003eC value of the added C\u003csub\u003e4\u003c/sub\u003e litter was always more positive than δ\u003csup\u003e13\u003c/sup\u003eC value of CO\u003csub\u003e2\u003c/sub\u003e derived from soil organic matter at a given site.\u003c/p\u003e \u003cp\u003eSoil samples were incubated under oxic conditions in the dark at 23 \u003csup\u003eo\u003c/sup\u003eC for 571 d. Soil was kept in an open 50 mL centrifuge tube inside a glass jar (946 mL) sealed with a gas-tight aluminum lid with butyl septa for headspace gas purging and sampling. The jars were flushed with CO\u003csub\u003e2\u003c/sub\u003e-free air following periodic headspace sampling as described below, and CO\u003csub\u003e2\u003c/sub\u003e concentrations remained below 5000 ppm during the incubation. Soil moisture was monitored by recording the mass of each sample, and water was added every month before 179 d and every other month thereafter (due to the less frequent gas sampling) to replenish vapor lost during headspace flushing. To monitor instantaneous decomposition over time and to avoid CO\u003csub\u003e2\u003c/sub\u003e saturation in the jar, headspace gas was initially measured at 4 d and 11 d, every other week for another 140 d, and then every month after 179 d (for a total duration of 571 d). The CO\u003csub\u003e2\u003c/sub\u003e concentrations and their δ\u003csup\u003e13\u003c/sup\u003eC values were measured by a tunable diode laser absorption spectrometer (TDLAS, TGA200A, Campbell Scientific, Logan, UT) immediately prior to flushing the headspace \u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. Because jars remained sealed between headspace sampling events, we were able to quantify the entire cumulative production of CO\u003csub\u003e2\u003c/sub\u003e and its δ\u003csup\u003e13\u003c/sup\u003eC value from each replicate over the course of the experiment. The CO\u003csub\u003e2\u003c/sub\u003e production from soil was measured on samples with no addition of litter and lignin, and CO\u003csub\u003e2\u003c/sub\u003e from litter and \u003csup\u003e13\u003c/sup\u003eC\u003csub\u003eβ\u003c/sub\u003e-labeled lignin was calculated by two-source mixing models that used measurements from the litter\u0026thinsp;+\u0026thinsp;unlabeled lignin and litter\u0026thinsp;+\u0026thinsp;\u003csup\u003e13\u003c/sup\u003eC\u003csub\u003eβ\u003c/sub\u003e-labeled lignin treatments, respectively \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e (see SI for more details). The C decomposition from soil, litter and lignin were expressed as percentages of their initial C masses (41.9 mg for litter and 264 \u0026micro;g for the \u003csup\u003e13\u003c/sup\u003eC\u003csub\u003eβ\u003c/sub\u003e atom of the labeled lignin, and a variable amount for SOC; Supplementary Table S1).\u003c/p\u003e \u003cp\u003eWe also conducted an N addition experiment to test the effects of N availability on lignin and litter decomposition, using additional subsamples of the 0\u0026ndash;15 cm soils collected from the four sampling points described above. For this experiment, the subsamples amended with litter\u0026thinsp;+\u0026thinsp;unlabeled lignin or litter\u0026thinsp;+\u0026thinsp;\u003csup\u003e13\u003c/sup\u003eC\u003csub\u003eβ\u003c/sub\u003e-labeled lignin were also amended with NH\u003csub\u003e4\u003c/sub\u003eNO\u003csub\u003e3\u003c/sub\u003e at 50 mg N g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Note that the amount of N that we added is relatively high but comparable to the level of inorganic N often observed in agricultural fields after fertilization \u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e. Briefly, 51 mL of 0.0386 mol L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e NH\u003csub\u003e4\u003c/sub\u003eNO\u003csub\u003e3\u003c/sub\u003e was added to soil samples, and then more water was added as necessary to achieve field moisture capacity. Sample incubation and gas measurements were the same as described above and conducted over 18 months.\u003c/p\u003e \u003cp\u003e \u003cb\u003eField incubation experiments.\u003c/b\u003e The 0\u0026ndash;15 cm soils from all 16 sampling points at each site were used for field incubation (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Soil subsamples (4.5 g dry mass equivalent) were gently mixed with litter\u0026thinsp;+\u0026thinsp;unlabeled lignin or litter\u0026thinsp;+\u0026thinsp;\u003csup\u003e13\u003c/sup\u003eC\u003csub\u003eβ\u003c/sub\u003e-labeled lignin according to the mass ratios and substrate treatments described above. The soil\u0026thinsp;+\u0026thinsp;litter mixtures were then transferred to mesh bags (8 cm x 8 cm in size; 55 \u0026micro;m nylon screen), which allowed entry of fungal hyphae, bacteria, and soil microfauna while minimizing particle loss \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. The mesh bags were sealed with hot glue and shipped back to the sites of origin and buried at a depth of 0\u0026ndash;15 cm at the same locations where soils were initially sampled, and geo-referenced to facilitate retrieval. The mesh bags with litter\u0026thinsp;+\u0026thinsp;unlabeled lignin were buried at even-numbered sampling points for each site, and those with litter\u0026thinsp;+\u0026thinsp;\u003csup\u003e13\u003c/sup\u003eC\u003csub\u003eβ\u003c/sub\u003e-labeled lignin were buried at the odd-numbered sampling points. After approximately 1 y of field incubation, the mesh bags were retrieved by NEON staff, flash-frozen on dry ice, and shipped on ice to ISU. Some bags were damaged or could not be located in the field (31 out of 320 samples).\u003c/p\u003e \u003cp\u003eThe soil and litter mixture was subsampled from each mesh bag, and then air-dried and finely ground for analysis of C concentrations and δ\u003csup\u003e13\u003c/sup\u003eC at the UC Davis Stable Isotope Facility using an elemental analyzer (Elementar Analysensysteme GmbH, Hanau, Germany) and continuous flow isotope ratio mass spectrometer (Sercon Ltd., Cheshire, UK). Lignin C remaining after the 1-y field incubation was calculated by multiplying \u003cem\u003ef\u003c/em\u003e\u003csub\u003elignin\u003c/sub\u003e calculated based on two-source mixing model (see details in SI) by the total C concentration in samples from the soil\u0026thinsp;+\u0026thinsp;litter\u0026thinsp;+\u0026thinsp;\u003csup\u003e13\u003c/sup\u003eC\u003csub\u003eβ\u003c/sub\u003e-labeled lignin treatment, with corrections accounting for new C inputs as necessary based on measurements of the samples with unlabeled lignin (see details in SI).\u003c/p\u003e \u003cp\u003e \u003cb\u003eSoil inorganic N availability.\u003c/b\u003e We measured ammonium (NH\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e+\u003c/sup\u003e) and nitrate (NO\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e) in additional replicate soil\u0026thinsp;+\u0026thinsp;litter mixture samples (10:1 mass ratio of soil to litter) from all soils used in the lab incubation after 1, 9, and 18 months. Briefly, 10 g soil mixed with 1 g litter was placed in a 50 mL centrifuge tube, loosely covered, and then incubated at 23\u0026deg;C in the dark after adjusting soil moisture to field capacity. Water was periodically added to soil samples to replace vapor loss, measured gravimetrically. Soil (~\u0026thinsp;2g) was subsampled from each centrifuge tube and extracted with 2 M potassium chloride at each timepoint. The soil solution was analyzed by microplate colorimetry for NH\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e+\u003c/sup\u003e-N \u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. The NO\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e-N was analyzed by microplate colorimetry \u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e or for the 9-month samples, second-derivative spectroscopy \u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e; these methods agreed almost perfectly on a subset of samples (slope\u0026thinsp;=\u0026thinsp;0.95, R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.97). Net N mineralization was calculated as the difference in inorganic N between sets of sampling points (9-month vs. 1-month; 18-month vs. 9-month; 18-month vs. 1-month).\u003c/p\u003e \u003cp\u003e \u003cb\u003eSoil geochemical analysis.\u003c/b\u003e Most physical and geochemical measurements were made on soils from all of the sampling points used for the field and lab incubations, except for particle size and 0.5 M HCl extractions, which were done for the four sampling points per site used for laboratory incubation. Physical and geochemical measurements included soil pH, particle size fractions, 0.5 M HCl-extractable Fe(II) and Fe(III), ammonium oxalate-extracted metals (Al, Fe, Mn), and citrate dithionite-extracted metals (Al, Fe, Mn and Ca). Some of these data were presented previously in a manuscript describing relationships between soil properties and particulate and mineral-associated organic matter fractions of these soils \u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. Field-moist soil subsamples were measured for pH in 1:1 slurries of soil and deionized water. Air-dried subsamples were used to measure particle size (sand, silt and clay) by sieving and sedimentation following aggregate dispersion with sodium hexametaphosphate \u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. Field-moist subsamples were extracted with 0.5 M hydrochloric acid (HCl) to measure dissolved and adsorbed Fe(II) as well as dissolved, organically-complexed Fe(III), and a highly reactive fraction of Fe(III) minerals \u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e. Concentrations of Fe(II) and Fe(III) were measured colorimetrically \u003csup\u003e\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e and summed as Fe\u003csub\u003eHCl\u003c/sub\u003e. Additional air-dried subsamples were extracted with acid ammonium oxalate in the dark at pH 3 to measure organo-metal complexes and short-range-ordered (SRO) phases of Al, Fe, and Mn (denoted Al\u003csub\u003eox\u003c/sub\u003e, Fe\u003csub\u003eox\u003c/sub\u003e, Mn\u003csub\u003eox\u003c/sub\u003e), and with sodium citrate dithionite to measure the crystalline and SRO phases of Fe (Fe\u003csub\u003ecd\u003c/sub\u003e) as well co-occurring Al, Mn, and Ca (Al\u003csub\u003ecd\u003c/sub\u003e, Mn\u003csub\u003ecd\u003c/sub\u003e, and Ca\u003csub\u003ecd\u003c/sub\u003e) \u003csup\u003e\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e. Metals were analyzed via inductively coupled plasma optical emission spectrometry (PerkinElmer Optima 5300 DV, Waltham, MA). Extraction of Al and Mn by oxalate and citrate-dithionite were very similar, so we only report Al\u003csub\u003eox\u003c/sub\u003e and Mn\u003csub\u003ecd\u003c/sub\u003e. The difference between Fe\u003csub\u003ecd\u003c/sub\u003e and Fe\u003csub\u003eox\u003c/sub\u003e represents crystalline phases (Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e). We interpret Mn\u003csub\u003ecd\u003c/sub\u003e as including exchangeable Mn, organo-metal complexes, and poorly crystalline phases. We interpret Ca\u003csub\u003ecd\u003c/sub\u003e as a measure of exchangeable Ca and Ca in organo-Fe associations \u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003cb\u003eMicrobial analysis.\u003c/b\u003e DNA was extracted from soils for ITS amplicon sequencing and quantitative PCR of ITS and 16S regions. Each of the four soils per site used for lab incubation was subsampled for DNA extraction at the beginning of the incubation, and additional replicates were extracted after 9 and 14 months. The incubated replicates used for DNA extraction were prepared similarly to the replicates used for CO\u003csub\u003e2\u003c/sub\u003e analyses, and were amended with \u003cem\u003eA. gerardi\u003c/em\u003e litter in a 1:10 mass ratio of litter to soil. The field-incubated soils corresponding to the same four sampling points for each site used in the lab incubation were also extracted for DNA, totaling 548 samples overall (156 soils \u0026times; 3 time points for lab incubation and 80 soils for field incubation). Soils were stored at -80\u0026deg;C before DNA extraction from 250 mg subsamples using the MagAttract PowerSoil DNA EP Kit (Qiagen, USA) on an Eppendorf epMotion 5075 liquid handling robot (Eppendorf North America, USA). Concentrations of DNA were measured using a Quant-iT\u0026trade; dsDNA high sensitivity Assay Kit (Invitrogen, USA) to standardize DNA masses for sequencing. Samples were diluted to 10 ng DNA \u0026micro;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e prior to sequencing; samples with concentration\u0026thinsp;\u0026lt;\u0026thinsp;10 ng DNA \u0026micro;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e were submitted directly. The ITS1 region of the ITS rRNA gene was amplified using the primer sets ITS1f (CTTGGTCATTTAGAGGAAGTAA) and ITS2 (GCTGCGTTCTTCATCGATGC), with PCR conditions as follows: 1 min at 94\u0026deg;C, followed by 35 cycles of 30 s at 94\u0026deg;C, 30 s at 52\u0026deg;C and 30 s at 68\u0026deg;C, and 10 min at 68\u0026deg;C. Fungal ITS rRNA gene amplicon sequencing was performed on the Illumina Miseq platform at Argonne National Laboratory with library preparation using the Miseq Reagent Kit V2 (Illumina, USA), producing 2 \u0026times; 250-bp reads. The DNA sequencing data are available at National Center for Biotechnology Information (NCBI) Sequence Read Archive PRJNA808104.\u003c/p\u003e \u003cp\u003eQuantitative real-time PCR was performed on a CFX96\u0026trade; real-time system coupled to a C1000\u0026trade; thermal cycler (Bio-Rad, USA) to assess the quantity of 16S rRNA and ITS genes. Each sample was prepared using 10 \u0026micro;L of SsoFast EvaGreen Supermix, 0.6 \u0026micro;L of each primer, 2 \u0026micro;L of diluted DNA sample, and nuclease-free water to a final volume of 20 \u0026micro;L. Bacterial 16S rRNA genes were amplified using the primer sets 1055YF(ATGGYTGTCGTCAGCT) and 1392R (ACGGGCGGTGTGTAC) and the following PCR conditions: 2 min at 50\u0026deg;C and 10 min at 95\u0026deg;C, followed by 40 cycles of 15 s at 95\u0026deg;C and 1 min at 58\u0026deg;C \u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e\u003c/sup\u003e. Fungal ITS rRNA genes were amplified using the primer sets ITS1F_KYO1 (CTHGGTCATTTAGAGGAASTAA) and ITS2_KYO2 (TTYRCTRCGTTCTTCATC) \u003csup\u003e\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e and the following PCR conditions: 2 min at 50\u0026deg;C and 2 min at 95\u0026deg;C, followed by 40 cycles of 30 s at 95\u0026deg;C, 30 s at 55\u0026deg;C and 60 sec at 72\u0026deg;C, and 10 min at 72\u0026deg;C. Standard curves for 16S rRNA and ITS rRNA genes were constructed using serial 10-fold dilutions from 10\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e to 10\u003csup\u003e\u0026minus;\u0026thinsp;8\u003c/sup\u003e of known concentrations of synthesized oligonucleotides (Integrated DNA Technologies, USA).\u003c/p\u003e \u003cp\u003e \u003cb\u003eBioinformatics.\u003c/b\u003e We used the Divisive Amplicon Denoising Algorithm 2 (DADA2) pipeline \u003csup\u003e\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e to process the ITS sequencing data in R statistical software version 3.6.1 \u003csup\u003e70\u003c/sup\u003e. We excluded samples with small numbers of reads (\u0026le;\u0026thinsp;900 sequences), including 27, 78, and 1 samples collected after 0-, 9-, and 14-months of the lab incubation, respectively. All functions were run using default parameters suggested by the DADA2 pipeline tutorial. The end product included an amplicon sequence variant (ASV) table recording the number of times each exact ASV was observed in each sample, along with a taxa table recording taxonomy assigned to the ASVs from kingdom to species levels, using the naive Bayesian classifier algorithm and the UNITE ITS database version 10.05.2021. Most ASVs had 251\u0026ndash;336 bp, falling within the commonly amplified ITS1 length of 200\u0026ndash;600 bp. Next, we trimmed the ASV tables using the \u0026ldquo;phyloseq\u0026rdquo; package (McMurdie and Holmes, 2013) in R. ASVs with \u0026lt;\u0026thinsp;10 sequences, i.e., rare ASVs, across all samples were removed. Before trimming, there were 22154 total ASVs and 3118076 total sequences across 442 samples; afterwards, there were 15583 total ASVs and 3085446 total sequences. After removing rare ASVs, there were 4 to 126 ASVs (mean\u0026thinsp;=\u0026thinsp;55) and 441 to 17234 sequences per sample (mean\u0026thinsp;=\u0026thinsp;6981). Rarefaction curves suggested that sequencing depths were adequate for all samples (data not shown).\u003c/p\u003e \u003cp\u003e \u003cb\u003eStatistical Analysis.\u003c/b\u003e For the lab incubation, we explored temporal trends in instantaneous C decomposition rate from each C source at each site and in lignin C decomposition rate for each individual sampling point (Supplementary Fig. S2), using generalized additive mixed models (GAMMs), including an autoregressive error term to account for temporal autocorrelation, using the \u0026ldquo;mgcv\u0026rdquo; package \u003csup\u003e\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e\u003c/sup\u003e version 1.8.28 in R 3.6.1. Pairwise correlations between cumulative C decomposition over 6, 12, and 18 months (lignin, litter, soil and field lignin decomposition) and biogeochemical predictors were tested by Pearson correlation. The biogeochemical predictors included several categories, which we define as follows (1) climatic: MAT and MAP; (2) N-related: bulk N, bulk C/N, NH\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e+\u003c/sup\u003e-N and NO\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e-N after 1-, 9- and 18-month incubations; (3) geochemical: soil pH, silt\u0026thinsp;+\u0026thinsp;clay, Al\u003csub\u003eox\u003c/sub\u003e, Fe\u003csub\u003eox\u003c/sub\u003e, Fe\u003csub\u003ecd\u0026minus;ox\u003c/sub\u003e, Fe\u003csub\u003eHCl\u003c/sub\u003e, Mn\u003csub\u003ecd\u003c/sub\u003e, Ca\u003csub\u003ecd\u003c/sub\u003e; (4) microbial: fungal composition, fungal Chao1 richness, fungal quantity, bacterial quantity, and fungal-to-bacterial ratio (Supplementary Table S2).\u003c/p\u003e \u003cp\u003eIn the microbial predictors, fungal composition was represented by the first (PC1) or second (PC2) axis of a principal coordinate analysis of ITS sequencing data on soils subsampled from the lab incubation at 14 months, conducted in the \u0026ldquo;vegan\u0026rdquo; package. The species-level abundance table (rather than the ASV table) was used to calculate Hellinger distances among samples before the analysis to alleviate the issue of a sparse matrix with many zero values \u003csup\u003e\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. The PC2 of fungal species composition was significantly (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) correlated with cumulative lignin and litter decomposition in the lab incubation and thus used as fungal composition predictor. Similarly, the PC1 of fungal species composition was significantly (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) correlated with cumulative SOC decomposition. Overall fungal composition changed little with time during the lab incubation (Supplementary Fig. S5b). Therefore, for subsequent statistical analyses we used the ITS data from samples collected after 14 months of incubation, because only one sample from this time point was excluded from analyses because of low read counts. Fungal richness was represented by the residual of ASV Chao1 index regressed on the square root of the number of total sequences within a sample, a method that accounts for differences in sequencing depth among samples \u003csup\u003e\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e\u003c/sup\u003e. We used copy numbers of ITS and 16S rRNA genes in the initial soil samples (1 g dry mass equivalent) as indices of fungal and bacterial quantity in our statistical models. Although fungal and bacterial quantities changed throughout the incubation (Supplementary Fig. S5a), including data from 9 and 14 months did not improve model performance. Fungal-to-bacterial ratio was calculated as fungal quantity divided by bacterial quantity.\u003c/p\u003e \u003cp\u003eWe also investigated whether putative lignin-degrading fungal organisms, i.e., white-, brown-, and soft-rot fungi identified in the FUNGuild database \u003csup\u003e\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e\u003c/sup\u003e, were linked to lignin decomposition. Similarly, we used PC1 or PC2 of a principal coordinate analysis based on Hellinger distances among samples from the lab incubation at 14 months calculated using only the ASV abundances of the identified \u0026ldquo;rot\u0026rdquo; fungi to represent their overall composition. We summed relative abundance of the identified \u0026ldquo;rot\u0026rdquo; fungi for each sample. Pearson correlations were then performed among lab lignin decomposition and overall composition and summed abundance of the \u0026ldquo;rot\u0026rdquo; fungi. We also used Pearson correlations to examine relationships among lab lignin decomposition and individual fungal genera occurring in more than 10 samples.\u003c/p\u003e \u003cp\u003eWe further used linear mixed models (LMMs) and random forest models (RFMs) to identify important predictors for cumulative C decomposition (lignin, litter, soil, and field lignin) variables. We included the above-mentioned climatic, N-related, geochemical, and microbial predictors in models of the laboratory incubation decomposition data. Inorganic N predictors from three timepoints could explain some variation in lab litter decomposition in the RFM but including these predictors did not improve model performance or change variable importance of other key predictors. Thus, inorganic N predictors were not retained in the final models, and we conducted the above-mentioned N addition experiment to specifically test the effects of inorganic N on lignin and litter decomposition. For statistical models of field lignin decomposition, we selected the climatic and geochemical predictors described above. We first fit the models including all categories of predictors and found that microbial predictors, silt\u0026thinsp;+\u0026thinsp;clay, and Fe\u003csub\u003eHCl\u003c/sub\u003e were not important predictors of field lignin decomposition. Therefore, we re-fit the models excluding these candidate predictors because these data were collected only for the field samples from the locations corresponding to the lab incubation. Inorganic N variables in soil\u0026thinsp;+\u0026thinsp;litter mixtures were not measured for field lignin decomposition.\u003c/p\u003e \u003cp\u003eIn the LMM, homoscedasticity and normality assumptions were met by raw data, except for lab lignin decomposition, which was log10 transformed. To estimate predictor importance, all variables were standardized to a mean of zero and a standard deviation of one to account for magnitude difference. All predictor variables were used as fixed effects and site was included as a random intercept to account for possible intra-site dependence in the LMMs. Adding sampling location as an additional random effect to account for correlations between 0\u0026ndash;15 and 15\u0026ndash;30 cm samples did not improve model performance. Some candidate predictors were excluded from initial models because of weak pairwise correlations with response variables (usually r\u0026thinsp;\u0026lt;\u0026thinsp;0.10), and/or moderate-to-strong collinearities with other predictors (usually r\u0026thinsp;\u0026gt;\u0026thinsp;0.50). Some predictors were further removed from final models through comparison of Akaike Information Criterion (AIC) values of nested models using stepwise backward selection. All predictors in the final models exhibited variance inflation factor values\u0026thinsp;\u0026lt;\u0026thinsp;3 and correlation coefficients\u0026thinsp;\u0026lt;\u0026thinsp;0.70 or \u0026gt; -0.70, implying that collinearity was acceptable. The relative contributions of fixed effects were determined by standardized regression coefficient estimates, and their significance was tested by the Wald chi-square test. LMM performance was evaluated by R\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e representing variance explained by only the fixed effects and by the model, respectively. The LMM analyses were conducted with the \u0026ldquo;lme4\u0026rdquo; package \u003csup\u003e\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWe used random forest models (RFMs) to explore possible nonlinear relationships among predictors and C decomposition variables (Breiman 2001). Variables were not standardized for an easier interpretation of the RFM partial dependence plot, which showed the marginal effect of each predictor on the predicted response variable. Lab lignin decomposition was log10 transformed and the same predictors as in the initial LMMs were included in the initial RFMs. Unimportant predictors were removed from models with Z-score\u0026thinsp;\u0026lt;\u0026thinsp;5 in the \u0026ldquo;Boruta\u0026rdquo; package. Some predictors were further removed from final models based on root mean square error (RMSE) and R\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e using repeated (N\u0026thinsp;=\u0026thinsp;100) cross validation. The RFMs were not overfit as indicated by an overfitting ratio\u0026thinsp;\u0026gt;\u0026thinsp;10 in the \u0026ldquo;rfUtilities\u0026rdquo; package. RFM was applied with 1,000 trees, with other options sticking to default parameters in the \u0026ldquo;randomForest\u0026rdquo; package \u003csup\u003e\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e\u003c/sup\u003e. The RFM performance was evaluated by R\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, and model overfitting was examined using the \u0026ldquo;rfUtilities\u0026rdquo; package. Variable importance was assessed using increase of mean squared error (%IncMSE) when a given variable is randomly permuted; a larger increase in MSE illustrates greater importance of the permuted variable. All statistical analyses and plotting were performed in R statistical software version 3.6.1 \u003csup\u003e70\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eData availability:\u003c/strong\u003e The data from this study will be available from the Environmental Data Initiative Data Portal upon acceptance.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank all of the NEON staff who contributed to field sampling, and A. Mirabito, S. Tsui, L. James, A. Boyer and H. Craven for lab assistance. This work was funded in part by National Science Foundation grant 1802745 (SJH, SRW, AH, CL) and Office of Biological and Environmental Research of the U.S Department of Energy Great Lakes Bioenergy Research Center grants DE-FC02-07ER64494, DE-SC0012742.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eS.J.H. and S.R.W. conceived and designed this study. K.E.H. and V.I.T. conducted lignin syntheses. W.H. and W.Y. performed research. W.Y., E.R. and J.Y. conduct microbial analysis. W.Y. and W.H. analyzed the data. W.H., W.Y. and S.J.H. wrote the manuscript. S.R.W., B.Y., K.E.H., C.L. and A.C.H. provided suggestions for substantial revisions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdditional information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplementary information 1: \u003cstrong\u003ePreparation of litter and lignin mixture\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplementary information 2: \u003cstrong\u003ePartitioning of decomposed C sources in the lab incubation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplementary information 3: \u003cstrong\u003eLignin C decomposition in the field incubation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplementary Figs. S1 to S11\u003c/p\u003e\n\u003cp\u003eSupplementary Tables S1 to S3\u003c/p\u003e\n\u003cp\u003eReferences\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests:\u003c/strong\u003e Authors declare that they have no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eMelillo, J. M., Aber, J. D. \u0026amp; Muratore, J. F. Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology \u003cb\u003e63\u003c/b\u003e, 621\u0026ndash;626 (1982).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAdair, E. C. \u003cem\u003eet al.\u003c/em\u003e Simple three-pool model accurately describes patterns of long-term litter decomposition in diverse climates. Glob. Chang. Biol. \u003cb\u003e14\u003c/b\u003e, 2636\u0026ndash;2660 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBollag, J.-M., Dec, J. \u0026amp; Huang, P. M. Formation mechanisms of complex organic structures in soil habitats. in \u003cem\u003eAdvances in Agronomy\u003c/em\u003e (ed. Sparks, D. L.) vol.\u0026nbsp;63 237\u0026ndash;266 (Academic Press, MA, 1997).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, J. \u003cem\u003eet al.\u003c/em\u003e A keystone microbial enzyme for nitrogen control of soil carbon storage. Sci. Adv. \u003cb\u003e4\u003c/b\u003e, eaaq1689 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHall, S. J., Huang, W., Timokhin, V. I. \u0026amp; Hammel, K. E. Lignin lags, leads, or limits the decomposition of litter and soil organic carbon. Ecology \u003cb\u003e101\u003c/b\u003e, e03113 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTalbot, J. M. \u0026amp; Treseder, K. K. Interactions among lignin, cellulose, and nitrogen drive litter chemistry\u0026ndash;decay relationships. Ecology \u003cb\u003e93\u003c/b\u003e, 345\u0026ndash;354 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThevenot, M., Dignac, M.-F. \u0026amp; Rumpel, C. Fate of lignins in soils: A review. Soil Biol. Biochem. \u003cb\u003e42\u003c/b\u003e, 1200\u0026ndash;1211 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCotrufo, M. F., Wallenstein, M. D., Boot, C. M., Denef, K. \u0026amp; Paul, E. The Microbial Efficiency-Matrix Stabilization (MEMS) framework integrates plant litter decomposition with soil organic matter stabilization: do labile plant inputs form stable soil organic matter? Glob. Chang. Biol. \u003cb\u003e19\u003c/b\u003e, 988\u0026ndash;995 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHall, S. J., Ye, C., Weintraub, S. R. \u0026amp; Hockaday, W. C. Molecular trade-offs in soil organic carbon composition at continental scale. Nat. Geosci. \u003cb\u003e13\u003c/b\u003e, 687\u0026ndash;692 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRasmussen, C., Matsuyama, N., Dahlgren, R. A., Southard, R. J. \u0026amp; Brauer, N. Soil genesis and mineral transformation across an environmental gradient on andesitic lahar. Soil Sci. Soc. Am. J. \u003cb\u003e71\u003c/b\u003e, 225\u0026ndash;237 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDoetterl, S. \u003cem\u003eet al.\u003c/em\u003e Soil carbon storage controlled by interactions between geochemistry and climate. Nat. Geosci. \u003cb\u003e8\u003c/b\u003e, 780\u0026ndash;783 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoorhead, D. L. \u0026amp; Sinsabaugh, R. L. A theoretical model of litter decay and microbial interaction. Ecol. Monogr. \u003cb\u003e76\u003c/b\u003e, 151\u0026ndash;174 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKleber, M. \u003cem\u003eet al.\u003c/em\u003e Chapter One - Mineral\u0026ndash;organic associations: Formation, properties, and relevance in soil environments. in \u003cem\u003eAdvances in Agronomy\u003c/em\u003e (ed. Sparks, D. L.) vol.\u0026nbsp;130 1\u0026ndash;140 (Academic Press, MA, 2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaiser, K., Guggenberger, G. \u0026amp; Haumaier, L. Changes in dissolved lignin-derived phenols, neutral sugars, uronic acids, and amino sugars with depth in forested Haplic Arenosols and Rendzic Leptosols. Biogeochemistry \u003cb\u003e70\u003c/b\u003e, 135\u0026ndash;151 (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCoward, E. K., Ohno, T. \u0026amp; Plante, A. F. Adsorption and molecular fractionation of dissolved organic matter on iron-bearing mineral matrices of varying crystallinity. Environ. Sci. Technol. \u003cb\u003e52\u003c/b\u003e, 1036\u0026ndash;1044 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, W. \u003cem\u003eet al.\u003c/em\u003e Enrichment of lignin-derived carbon in mineral-associated soil organic matter. Environ. Sci. Technol. \u003cb\u003e53\u003c/b\u003e, 7522\u0026ndash;7531 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, H., Santos, F., Butler, K. \u0026amp; Herndon, E. A critical review on the multiple roles of manganese in stabilizing and destabilizing soil organic matter. Environ. Sci. Technol. \u003cb\u003e55\u003c/b\u003e, 12136\u0026ndash;12152 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeiluweit, M. \u003cem\u003eet al.\u003c/em\u003e Long-term litter decomposition controlled by manganese redox cycling. PNAS \u003cb\u003e112\u003c/b\u003e, E5253\u0026ndash;E5260 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRowley, M. C., Grand, S. \u0026amp; Verrecchia, \u0026Eacute;. P. Calcium-mediated stabilisation of soil organic carbon. Biogeochemistry \u003cb\u003e137\u003c/b\u003e, 27\u0026ndash;49 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLovett, G. M., Arthur, M. A. \u0026amp; Crowley, K. F. Effects of calcium on the rate and extent of litter decomposition in a northern hardwood forest. Ecosystems \u003cb\u003e19\u003c/b\u003e, 87\u0026ndash;97 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLouca, S. \u003cem\u003eet al.\u003c/em\u003e Function and functional redundancy in microbial systems. Nat. Ecol. Evol. \u003cb\u003e2\u003c/b\u003e, 936\u0026ndash;943 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBugg, T. D., Ahmad, M., Hardiman, E. M. \u0026amp; Singh, R. The emerging role for bacteria in lignin degradation and bio-product formation. Curr. Opin. Biotechnol. \u003cb\u003e22\u003c/b\u003e, 394\u0026ndash;400 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHammel, K. E. Fungal degradation of lignin. in \u003cem\u003eDriven by Nature: Plant Litter Quality and Decomposition\u003c/em\u003e (eds. Cadisch, G. \u0026amp; Giller, K. E.) 33\u0026ndash;46 (CAB International, Wallingford, 1997).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJanusz, G. \u003cem\u003eet al.\u003c/em\u003e Lignin degradation: microorganisms, enzymes involved, genomes analysis and evolution. FEMS Microbiol. Rev. \u003cb\u003e41\u003c/b\u003e, 941\u0026ndash;962 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchimel, J. \u0026amp; Schaeffer, S. M. Microbial control over carbon cycling in soil. Front. Microbiol. \u003cb\u003e3\u003c/b\u003e, Article 348 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOsono, T. Decomposing ability of diverse litter-decomposer macrofungi in subtropical, temperate, and subalpine forests. J. For. Res. \u003cb\u003e20\u003c/b\u003e, 272\u0026ndash;280 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerg, B. \u0026amp; Staaf, H. Decomposition rate and chemical changes of scots pine needle litter. II. Influence of chemical composition. Ecol. Bull. \u003cb\u003e32\u003c/b\u003e, 373\u0026ndash;390 (1980).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKallenbach, C. M., Frey, S. D. \u0026amp; Grandy, A. S. Direct evidence for microbial-derived soil organic matter formation and its ecophysiological controls. Nat. Commun. \u003cb\u003e7\u003c/b\u003e, 13630 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSluiter, J. B., Ruiz, R. O., Scarlata, C. J., Sluiter, A. D. \u0026amp; Templeton, D. W. Compositional analysis of lignocellulosic feedstocks. 1. Review and description of methods. J. Agric. Food Chem. \u003cb\u003e58\u003c/b\u003e, 9043\u0026ndash;9053 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHernes, P. J., Kaiser, K., Dyda, R. Y. \u0026amp; Cerli, C. Molecular trickery in soil organic matter: Hidden lignin. Environ. Sci. Technol. \u003cb\u003e47\u003c/b\u003e, 9077\u0026ndash;9085 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaider, K. \u0026amp; Martin, J. P. Decomposition in soil of specifically \u003csup\u003e14\u003c/sup\u003eC-labeled model and cornstalk lignins and coniferyl alcohol over two years as influenced by drying, rewetting, and additions of an available C substrate. Soil Biol. Biochem. \u003cb\u003e13\u003c/b\u003e, 447\u0026ndash;450 (1981).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHall, S. J. \u0026amp; Thompson, A. What do relationships between extractable metals and soil organic carbon concentrations mean? Soil Sci. Soc. Am. J. \u003cb\u003e86\u003c/b\u003e, 195\u0026ndash;208 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, W., Weintraub, S. R. \u0026amp; Hall, S. J. Climatic and geochemical controls on soil carbon at the continental scale: interactions and thresholds. \u003cem\u003eGlobal Biogeochem. Cycles.\u003c/em\u003e \u003cb\u003e35\u003c/b\u003e, e2020GB006781 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, C., Hall, S. J., Coward, E. \u0026amp; Thompson, A. Iron-mediated organic matter decomposition in humid soils can counteract protection. Nat. Commun. \u003cb\u003e11\u003c/b\u003e, 2255 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHofrichter, M. Review: lignin conversion by manganese peroxidase (MnP). Enzyme Microb. Technol. \u003cb\u003e30\u003c/b\u003e, 454\u0026ndash;466 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrossman, J. J., Cavender-Bares, J. \u0026amp; Hobbie, S. E. Functional diversity of leaf litter mixtures slows decomposition of labile but not recalcitrant carbon over two years. Ecol. Monogr. \u003cb\u003e90\u003c/b\u003e, e01407 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerg, B. \u003cem\u003eet al.\u003c/em\u003e Maximum decomposition limits of forest litter types: a synthesis. Can. J. Bot. \u003cb\u003e74\u003c/b\u003e, 659\u0026ndash;672 (1996).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKleber, M. \u003cem\u003eet al.\u003c/em\u003e Dynamic interactions at the mineral\u0026ndash;organic matter interface. Nat. Rev. Earth Environ. \u003cb\u003e2\u003c/b\u003e, 402\u0026ndash;421 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrankland, J. C. Fungal decomposition of bracken petioles. J. Ecol. \u003cb\u003e57\u003c/b\u003e, 25\u0026ndash;36 (1969).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZare-Maivan, H. \u0026amp; Shearer, C. A. Extracellular enzyme production and cell wall degradation by freshwater lignicolous fungi. Mycologia \u003cb\u003e80\u003c/b\u003e, 365\u0026ndash;375 (1988).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFukasawa, Y. \u0026amp; Matsukura, K. Decay stages of wood and associated fungal communities characterise diversity\u0026ndash;decomposition relationships. Sci. Rep. \u003cb\u003e11\u003c/b\u003e, 8972 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTalbot, J. M. \u003cem\u003eet al.\u003c/em\u003e Endemism and functional convergence across the North American soil mycobiome. PNAS \u003cb\u003e111\u003c/b\u003e, 6341\u0026ndash;6346 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoer, W. de, Folman, L. B., Summerbell, R. C. \u0026amp; Boddy, L. Living in a fungal world: impact of fungi on soil bacterial niche development. FEMS Microbiol. Rev. \u003cb\u003e29\u003c/b\u003e, 795\u0026ndash;811 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchimel, J. P. Ecosystem consequences of microbial diversity and community structure. in \u003cem\u003eArctic and alpine biodiversity\u003c/em\u003e (eds. Chapin III, F. S. \u0026amp; Korner, C.) 239\u0026ndash;244 (Springer-Verlag, Berlin, 1995).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCleveland, C. C. \u003cem\u003eet al.\u003c/em\u003e Litter quality versus soil microbial community controls over decomposition: a quantitative analysis. Oecologia \u003cb\u003e174\u003c/b\u003e, 283\u0026ndash;294 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrickland, M. S., Lauber, C., Fierer, N. \u0026amp; Bradford, M. A. Testing the functional significance of microbial community composition. Ecology \u003cb\u003e90\u003c/b\u003e, 441\u0026ndash;451 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, Y.-R. \u003cem\u003eet al.\u003c/em\u003e New insights into the role of microbial community composition in driving soil respiration rates. Soil Biol. Biochem. \u003cb\u003e118\u003c/b\u003e, 35\u0026ndash;41 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKemmitt, S. J. \u003cem\u003eet al.\u003c/em\u003e Mineralization of native soil organic matter is not regulated by the size, activity or composition of the soil microbial biomass\u0026mdash;a new perspective. Soil Biol. Biochem. \u003cb\u003e40\u003c/b\u003e, 61\u0026ndash;73 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWaring, B. G. \u003cem\u003eet al.\u003c/em\u003e From pools to flow: The PROMISE framework for new insights on soil carbon cycling in a changing world. Glob. Chang. Biol. \u003cb\u003e26\u003c/b\u003e, 6631\u0026ndash;6643 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrickland, M. S., Keiser, A. D. \u0026amp; Bradford, M. A. Climate history shapes contemporary leaf litter decomposition. Biogeochemistry \u003cb\u003e122\u003c/b\u003e, 165\u0026ndash;174 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlassman, S. I. \u003cem\u003eet al.\u003c/em\u003e Decomposition responses to climate depend on microbial community composition. PNAS \u003cb\u003e115\u003c/b\u003e, 11994\u0026ndash;11999 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBradford, M. Thermal adaptation of decomposer communities in warming soils. Front. Microbiol. \u003cb\u003e4\u003c/b\u003e, Article 333 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, W., Huang, W., Weintraub-Leff, S. R. \u0026amp; Hall, S. J. Where and why do particulate organic matter (POM) and mineral-associated organic matter (MAOM) differ among diverse soils? Soil Biol. Biochem. \u003cb\u003e172\u003c/b\u003e, 108756 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRasmussen, C. \u003cem\u003eet al.\u003c/em\u003e Beyond clay: towards an improved set of variables for predicting soil organic matter content. Biogeochemistry \u003cb\u003e137\u003c/b\u003e, 297\u0026ndash;306 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWagai, R., Kajiura, M. \u0026amp; Asano, M. Iron and aluminum association with microbially processed organic matter via meso-density aggregate formation across soils: organo-metallic glue hypothesis. SOIL \u003cb\u003e6\u003c/b\u003e, 597\u0026ndash;627 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeller, M., Schimel, D. S., Hargrove, W. W. \u0026amp; Hoffman, F. M. A continental strategy for the National Ecological Observatory Network. Front. Ecol. Environ. \u003cb\u003e6\u003c/b\u003e, 282\u0026ndash;284 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSanClements, M. \u003cem\u003eet al.\u003c/em\u003e Collaborating with NEON. BioScience \u003cb\u003e70\u003c/b\u003e, 107\u0026ndash;107 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHall, S. J., Huang, W. \u0026amp; Hammel, K. E. An optical method for carbon dioxide isotopes and mole fractions in small gas samples: Tracing microbial respiration from soil, litter, and lignin. Rapid Commun. Mass Spectrom. \u003cb\u003e31\u003c/b\u003e, 1938\u0026ndash;1946 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLawrence, N. C., Tenesaca, C. G., VanLoocke, A. \u0026amp; Hall, S. J. Nitrous oxide emissions from agricultural soils challenge climate sustainability in the US Corn Belt. PANS \u003cb\u003e118\u003c/b\u003e, e2112108118 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeatherburn, M. W. Phenol-hypochlorite reaction for determination of ammonia. Anal. Chem. \u003cb\u003e39\u003c/b\u003e, 971\u0026ndash;974 (1967).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDoane, T. A. \u0026amp; Horw\u0026aacute;th, W. R. Spectrophotometric determination of nitrate with a single reagent. Anal. Lett. \u003cb\u003e36\u003c/b\u003e, 2713\u0026ndash;2722 (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCrumpton, W. G., Isenhart, T. M. \u0026amp; Mitchell, P. D. Nitrate and organic N analyses with second-derivative spectroscopy. Limnol. Oceanogr. \u003cb\u003e37\u003c/b\u003e, 907\u0026ndash;913 (1992).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHall, S. J. \u0026amp; Silver, W. L. Reducing conditions, reactive metals, and their interactions can explain spatial patterns of surface soil carbon in a humid tropical forest. Biogeochemistry \u003cb\u003e125\u003c/b\u003e, 149\u0026ndash;165 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, W. \u0026amp; Hall, S. J. Optimized high-throughput methods for quantifying iron biogeochemical dynamics in soil. Geoderma \u003cb\u003e306\u003c/b\u003e, 67\u0026ndash;72 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoeppert, R. H. \u0026amp; Inskeep, W. P. Iron. in \u003cem\u003eMethods of Soil Analysis\u003c/em\u003e 639\u0026ndash;664 (John Wiley \u0026amp; Sons, Ltd, Hoboken, NJ, 1996).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHall, S. J. \u0026amp; Huang, W. Iron reduction: a mechanism for dynamic cycling of occluded cations in tropical forest soils? Biogeochemistry \u003cb\u003e136\u003c/b\u003e, 91\u0026ndash;102 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRitalahti, K. M. \u003cem\u003eet al.\u003c/em\u003e Quantitative PCR targeting 16S rRNA and reductive dehalogenase genes simultaneously monitors multiple \u003cem\u003eDehalococcoides\u003c/em\u003e strains. Appl. Environ. Microbiol. \u003cb\u003e72\u003c/b\u003e, 2765\u0026ndash;2774 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eToju, H., Tanabe, A. S., Yamamoto, S. \u0026amp; Sato, H. High-coverage ITS primers for the DNA-based identification of \u003cem\u003eAscomycetes\u003c/em\u003e and \u003cem\u003eBasidiomycetes\u003c/em\u003e in environmental samples. PLoS One \u003cb\u003e7\u003c/b\u003e, e40863 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCallahan, B. J. \u003cem\u003eet al.\u003c/em\u003e DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods \u003cb\u003e13\u003c/b\u003e, 581\u0026ndash;583 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR Core Team. \u003cem\u003eR: A language and environment for statistical computing\u003c/em\u003e. (R Foundation for Statistical Computing, Vienna, 2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWood, S. N. \u003cem\u003eGeneralized Additive Models: An Introduction with R, Second Edition\u003c/em\u003e. (Chapman and Hall/CRC, NY, 2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLegendre, P. \u0026amp; Gallagher, E. D. Ecologically meaningful transformations for ordination of species data. Oecologia \u003cb\u003e129\u003c/b\u003e, 271\u0026ndash;280 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTedersoo, L. \u003cem\u003eet al.\u003c/em\u003e Global diversity and geography of soil fungi. Science \u003cb\u003e346\u003c/b\u003e, 1256688 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNguyen, N. H. \u003cem\u003eet al.\u003c/em\u003e FUNGuild: An open annotation tool for parsing fungal community datasets by ecological guild. Fungal Ecol. \u003cb\u003e20\u003c/b\u003e, 241\u0026ndash;248 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBates, D., Maechler, M., Bolker, B. \u0026amp; Walker, S. Fitting linear mixed-effects models using lme4. Journal of Statistical Software \u003cb\u003e67\u003c/b\u003e, 1\u0026ndash;48 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreiman, L. Random forests. Machine Learning \u003cb\u003e45\u003c/b\u003e, 5\u0026ndash;32 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiaw, A. \u0026amp; Wiener, M. Classification and regression by random forest. R News \u003cb\u003e2\u003c/b\u003e, 18\u0026ndash;22 (2002).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-2086399/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-2086399/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eLignin is an abundant and complex plant polymer that may limit litter decomposition, yet lignin is sometimes a minor constituent of soil organic carbon (SOC). Accounting for geographic diversity in soil characteristics might reconcile this apparent contradiction. We tracked decomposition of a lignin/litter mixture across North American mineral soils using lab and field incubations. Cumulative lignin decomposition varied 18-fold among soils and was strongly correlated with bulk litter decomposition, but not SOC decomposition. Legacy climate predicted decomposition even in the lab. Impacts of nitrogen availability were minor compared with geochemical and microbial properties, which had contradictory relationships with lignin and SOC decomposition. Lignin decomposition increased with some metals and fungi, whereas SOC decomposition decreased with all metals and was weakly related with fungi. Soil properties differentially impact decomposition of lignin and litter vs. SOC across broad geographic scales, linking short-term decomposition to differences in organic matter among ecosystems.\u003c/p\u003e","manuscriptTitle":"Contrasting geochemical and fungal controls on decomposition of lignin and soil carbon at continental scale","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2022-09-23 18:45:56","doi":"10.21203/rs.3.rs-2086399/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"dc004185-21bf-4128-9d3e-ff607d817e83","owner":[],"postedDate":"September 23rd, 2022","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2023-04-20T07:10:44+00:00","versionOfRecord":{"articleIdentity":"rs-2086399","link":"https://doi.org/10.1038/s41467-023-37862-6","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2023-04-19 04:00:00","publishedOnDateReadable":"April 19th, 2023"},"versionCreatedAt":"2022-09-23 18:45:56","video":"","vorDoi":"10.1038/s41467-023-37862-6","vorDoiUrl":"https://doi.org/10.1038/s41467-023-37862-6","workflowStages":[]},"version":"v1","identity":"rs-2086399","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-2086399","identity":"rs-2086399","version":["v1"]},"buildId":"_2-kVJe1T_tPrBINL-cwx","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}