INTRODUCTION
Soils are a significant component of the Earth’s carbon (C) cycle (Eswaran et al. Reference Eswaran, Van Den Berg and Reich1993; Batjes Reference Batjes1996; Jobbágy and Jackson Reference Jobbágy and Jackson2000), yet a mechanistic understanding of what controls the turnover of this large C pool remains elusive. Soil organic C (SOC) stocks are primarily controlled by the balance of plant-derived C inputs and subsequent CO2 efflux due to microbial decomposition and root respiration (Davidson and Janssens Reference Davidson and Janssens2006). Microbial respiration of organic C accounts for roughly half of the total CO2 production from soils (Bond-Lamberty et al. Reference Bond-Lamberty, Wang and Gower2004), though this number varies with ecosystem type, temperature, and moisture (Subke et al. Reference Subke, Inglima and Cotrufo2006). The SOC used by microorganisms therefore has a significant impact on soil C cycling, influencing what SOC is cycled rapidly versus left to persist for centuries to millennia.
Radiocarbon (14C) is the gold standard for determining both the age and turnover rate of soil C, providing an invaluable metric for evaluating long-term C stability. Given the importance of microbial SOC cycling, many studies use laboratory soil incubations to measure the rate of heterotrophic respiration and the ∆14C of respired CO2 to assess C turnover utilization by microbes. While incubations provide an integrated assessment of microbial respiration and C turnover, soil sampling and preparation prior to incubation can result in artifacts due to the disruption of soil structure, roots, and microbial communities (Salomé et al. Reference Salomé, Nunan, Pouteau, Lerch and Chenu2010; Herbst et al. Reference Herbst, Tappe, Kummer and Vereecken2016; Schädel et al Reference Schädel, Beem-Miller, Aziz Rad, Crow, Hicks Pries, Ernakovich, Hoyt, Plante, Stoner, Treat and Sierra2020; Patel et al. Reference Patel, Bond-Lamberty, Jian, Morris, McKever, Norris, Zheng and Bailey2022). Comparisons between field-based and laboratory incubation studies show differences in gas flux rates (Williams et al. Reference Williams, Jarvis and Dixon1998; Risk et al. Reference Risk, Kellman, Beltrami and Diochon2008; Patel et al. Reference Patel, Bond-Lamberty, Jian, Morris, McKever, Norris, Zheng and Bailey2022) and younger respired C in the field (Phillips et al. Reference Phillips, McFarlane, Risk and Desai2013), suggesting that additional methods to assess microbial processes would be valuable.
To date, very few techniques other than laboratory incubations have been developed to specifically measure the ∆14C of organic C used by microbial communities. The only existing alternatives have relied on modifying the traditional chloroform fumigation extraction (CFE) approach–conducted by fumigating a soil with chloroform and then extracting the released biomolecules using a salt solution (Vance et al. Reference Vance, Brookes and Jenkinson1987). With CFE, the quantity of C is compared to a control extraction conducted without chloroform; the difference between the two is a measure of the total microbial biomass. Fearing that chloroform C contamination might render natural abundance 14C analysis impractical, Rumpel et al. (Reference Rumpel, Grootes and Kögel-Knabner2001) opted to rupture microbial cells using freeze-drying cycles rather than chloroform. However, Garnett et al. (Reference Garnett, Bol, Bardgett, Wanek, Bäumler and Richter2011) successfully used the traditional CFE protocol and found the chloroform C contamination was manageable, however their method requires a specialized vacuum system.
A more quantitative estimate of the age and turnover time of various soil organic pools is a key prerequisite to more accurate modeling of the stability of SOM under varying edaphic conditions. Here, we report on a new microbial biomass extraction method for 14C analysis, allowing for the empirical measurement of microbially assimilated C. The method is based on direct chloroform extraction which applies chloroform directly to the soil (Gregorich et al Reference Gregorich, Wen, Voroney and Kachanoski1990; Setia et al. Reference Setia, Lata Verma and Marschner2012; Slessarev et al. Reference Slessarev, Lin, Jiménez, Homyak, Chadwick, D’Antonio and Schimel2020). We compare the results of our 14C biomass extraction method to those of a traditional laboratory incubation from a soil profile to evaluate the utility of the method and future applications. Additionally, we evaluate the 14C blank contribution of our chloroform extraction protocol using a size series of 14C modern and fossil standards.
METHODS
Soil Sampling, Storage, and Bulk Soil Analysis
The soil samples used in this study were collected from the University of California Hopland Research and Extension Center in Hopland, CA in January 2022 (39.001º, -123.069º). The mean annual temperature and precipitation at the site are 15˚C and 940 mm/y, respectively, and the soil is classified as a Typic Haploxeralf on Cretaceous sandstone and shale (Foley et al. Reference Foley2022; Fossum et al. Reference Fossum, Estera-Molina, Yuan, Herman, Chu-Jacoby, Nico, Morrison, Pett-Ridge and Firestone2022). Samples were collected from a soil pit face at depth increments of 0–10 cm, 10–20 cm, 20–50 cm, and 50–100 cm. One aliquot of each sample was sealed in a bag and left at room temperature for one week until processing for laboratory incubations. A second aliquot of each sample was sealed in a bag and kept at 4ºC until use in microbial biomass extractions, approximately three months. Upon returning from the field, a subsample of bulk soil from each depth was air dried, sieved to 2 mm, and then ground in a ball mill. Triplicate samples of the ground bulk soil were sealed into quartz tubes for 14C and δ13C analysis, respectively.
Laboratory Soil Incubations
For each depth increment, three technical replicates were incubated. Between 90 and 200 g of soil was placed in a 32 oz jar after carefully removing visible roots with tweezers. Soil aggregates were intentionally left intact to minimize disturbance of the soil structure. After a 24 h pre-incubation at room temperature, the jars were flushed with > 4 times the headspace volume with certified CO2-free air and sealed. Incubations were conducted in triplicate from each depth increment and sampled periodically to determine headspace CO2 concentration via a LI-830 (LI-COR) infrared gas analyzer. After reaching ∼1% CO2, the headspace was transferred from each jar into a glass flask and immediately purified and graphitized for 14C analysis. The duration of incubation was dependent on the rate of CO2 respiration and ranged between 5 days for surface soils to 47 days for the deepest samples.
Microbial Biomass Extraction and Calculations
Microbial soil biomass was extracted and quantified based on a modified direct extraction method from Setia et al. (Reference Setia, Lata Verma and Marschner2012). Two technical replicate extractions were done from each soil depth to test the reproducibility of the method. To minimize C contamination, all glassware was acid washed and baked at 400ºC for 5 hr prior to use. 25 g of 2 mm sieved, field-moist soil was weighed into glass flasks along with 100 mL of Ultrapure water. For each sample, two soil slurries were prepared. 2.5 mL of ethanol-free chloroform (99%+ chloroform with ca 50 ppm amylene, Alfar Aesar, L14759) was added to one soil slurry, producing one “water” and one “chloroform” extract for each soil sample. The flasks were capped with glass stoppers and shaken in an orbital motion for 1 h at 140 RPM. The samples were vacuum filtered through pre-baked 0.7 μm glass fiber filters (400ºC for 5 hr), after which the filtrate was bubbled vigorously with Ultra-High Purity N2 (99.999%) for 30 min to remove any residual chloroform. N2 was introduced via pre-baked glass pipettes secured to a nitrogen evaporator. Extracts were finally filtered through a 0.2 μm polycarbonate filter to remove visible soil particles. For samples below 20 cm, extracts from three separate 25 g “water” or “chloroform” samples were pooled to recover sufficient C for 14C analysis, totaling 75 g of material.
A split of each sample was reserved for total organic carbon (TOC) analysis and the remainder was concentrated in an evaporative centrifuge. The concentrated biomass extracts were transferred to pre-baked (900ºC for 5 hr) 6 mm quartz tubes using 0.01 M HCl to remove any inorganic carbon, then dried to completion. CuO and Ag powder were added, and the sample tubes were loaded into 9 mm quartz tubes, evacuated, sealed, and combusted at 900ºC (Trumbore et al. Reference Trumbore, Xu, Santos, Czimczik, Beaupré, Pack, Hopkins, Stills, Lupascu, Ziolkowski, Schuur, Druffel and Trumbore2016). The quantity of the microbial biomass was calculated by subtracting the total organic C content of the water extract from the chloroform extract, and the Δ14C of the microbial biomass (MB) extract was calculated using (Garnett et al. Reference Garnett, Bol, Bardgett, Wanek, Bäumler and Richter2011):
$${\Delta ^{14}}\rm{C_{MB}} = \left( {{\Delta ^{14}}{C_C}*{C_C}-{\Delta ^{14}}{C_W}*{C_W}} \right)/\left( {{C_C}-{C_W}} \right)$$
where Δ14CC and Δ14CW refer to the measured 14C concentration of the chloroform and water, and CC and CW represent the mass of carbon in the chloroform and water extracts, respectively.
Blank Assessment and F14C Data Correction
To assess the C contamination (blank) introduced during the microbial biomass exactions, a size series of 14C-modern and -dead material (ANU sucrose and alanine, respectively) were processed in an identical fashion to the soil samples, in the range of 40 to 150 μg C. In total, 21 modern and 15 dead samples were analyzed in the size series. The size and fraction modern (F14C) of the blank were then determined using the methods and published R script from Sun et al. (Reference Sun, Meyer, Dolman, Winterfeld, Hefter, Dummann, McIntyre, Montluçon, Haghipour, Wacker, Gentz, van der Voort, Eglinton and Mollenhauer2020). Briefly, a Bayesian model was used to fit thousands of linear regression lines between the F14C and inverse of the sample size (1/μg C), allowing for the calculation of the F14C and size of the blank, as well as their associated uncertainties. The R script was run in R Studio version 4.1.2 (R Core Team 2021). The calculated blank was then used to correct the measured F14C of the water and chloroform extracts.
Sample Graphitization and Isotopic Analyses
Graphitization and accelerator mass spectrometry (AMS) measurements were conducted at the Center for Accelerator Mass Spectrometry (CAMS) at Lawrence Livermore National Laboratory. Bulk soil samples and microbial biomass extracts were prepared for graphitization through sealed-tube combustion at 900ºC in an evacuated quartz tube in the presence of CuO and Ag. The CO2 produced from sealed-tube combustion, as well as the headspace CO2 from the incubations, was purified and then reduced to graphite at 570ºC in the presence of iron powder and H2 (Vogel et al. Reference Vogel, Southon, Nelson and Brown1984). Samples were run on the model FN Van de Graaff AMS system at CAMS. During purification of the CO2, a split of each of the incubation and microbial biomass samples was taken and subsequently sent to the Stable Isotope Geosciences Facility at Texas A&M University for δ13C analysis on a Thermo Scientific MAT 253 Dual Inlet Stable Isotope Ratio Mass Spectrometer. Bulk soil samples were measured for % C and δ13C at the Center for Stable Isotope Biogeochemistry, University of California, Berkeley on a CHNOS Elemental Analyzer interfaced to an IsoPrime100 Isotope Ratio Mass Spectrometer. Measured radiocarbon values were corrected using offline δ13C values and reported as age-corrected Δ14C using the following equation and conventions from Stuiver and Polach, Reference Stuiver and Polach1977:
$$\Delta {}_\;^{14}C = \left( {{{{A_{SN\;}}{e^{\lambda \left( {1950 - x} \right)}}} \over {{A_{ON}}}} - 1} \right)*1000$$
where ASN is the normalized sample specific activity, AON is the normalized standard specific activity, λ is 1/8267 yr-1, and x is the year of measurement.
Statistical Analysis
Statistical analyses were conducted in R Studio version 4.1.2 (R Core Team 2021). Analysis of Variance (ANOVA) was used to test for significant differences in ∆14C value between incubation or biomass extraction at each depth.
RESULTS AND DISCUSSION
To assess the reliability and variance of the direct chloroform microbial biomass extraction, we compared ∆14C values of calculated microbial biomass from two replicate extractions to the ∆14C values of respired CO2 from three replicate incubations at each depth increment (Figure 1; Tables 1–2). Regardless of depth increment, the variance of ∆14C values from technical replicate soil incubations (n=3) was less than that of replicate biomass extractions (n=2), and the variability was larger at depth for both methods (Tables 1–2; Figure 1). In the upper 50 cm, the average Δ14C of respired CO2 was not significantly different than the Δ14C of the microbial biomass extract (p > 0.05) (Figure 1). Below 50 cm, the respired CO2 was significantly less depleted than the extracted biomass (p < 0.01). The average ∆14C of respired CO2 from the 0–10, 10–20, 20–50, and 50–100 cm depths was 6 ± 5, 17 ± 4, –3 ± 10, and –48 ± 17‰ (± SD, n=3) (Table 1; Figure 1), and the average ∆14C of extracted microbial biomass was 14 ± 17, 15 ± 10, 21 ± 22, and –220 ± 53‰ (± SD, n=2) (Table 2; Figure 1).
Figure 1 Average ∆14C value of bulk soil (n=3), respired CO2 from laboratory soil incubations (n=3), and soil microbial biomass from direct chloroform extraction (n=2) from a Hopland, CA annual grassland soil, sampled over four depth increments. Samples were collected from a soil pit face at depth increments of 0–10 cm, 10–20 cm, 20–50 cm, and 50–100 cm. Error bars indicate standard deviation of replicates.
Table 1 Incubation parameters and isotopic values of heterotrophically respired CO2 from triplicate laboratory incubations of soil collected at four depths from a grassland in Hopland, CA. Duration of the incubation was driven by respiration rate of individual samples. Incubations ran until sufficient CO2 had accumulated for radiocarbon analysis.
Table 2 Blank corrected measured radiocarbon values and measurement error of water and chloroform extracts and the calculated ∆14C of microbial biomass (Eq. 1) from a grassland soil collected at four depth increments in Hopland, CA.
We conducted a blank assessment by extracting a series of 14C-modern and -dead materials. From this blank assessment, we estimated that the biomass extraction protocol introduced 2.22 ± 0.40 μg C with a F14C value of 0.36 ± 0.08. Measured F14C values and AMS target sizes for the samples used in the blank assessment size series can be found in Supplemental Table 1.
Comparison of Biomass Extraction and Laboratory Incubation Methods
We found that both incubation and chloroform extraction methods of estimating microbial biomass C produced similar ∆14C results in the upper 50 cm soil increment (Table 1; Figure 1), indicating that for these surface soils, either method could be used to assess microbially used C. In contrast, the ∆14C values for soil collected from below 50 cm from the two methods diverge. It is possible that the soil sampling process and sample handling prior to incubation released fresh, labile C that otherwise would not have been accessible for decomposition (Salomé et al. Reference Salomé, Nunan, Pouteau, Lerch and Chenu2010; Herbst et al. Reference Herbst, Tappe, Kummer and Vereecken2016; Schädel et al Reference Schädel, Beem-Miller, Aziz Rad, Crow, Hicks Pries, Ernakovich, Hoyt, Plante, Stoner, Treat and Sierra2020; Patel et al. Reference Patel, Bond-Lamberty, Jian, Morris, McKever, Norris, Zheng and Bailey2022). Alternatively, the C sources used for respiration and assimilation may differ, which would result in diverging incubation and biomass values. Finally, it is also possible that the 14C depleted biomass values in the deeper soils may reflect non-living cell material that was liberated by the chloroform biomass extraction. This method should release all membrane-contained biomolecules from the soil, including microbial necromass and lipids, which previous reports suggest are the most persistent and 14C depleted compound class in soil (van der Voort et al. Reference van der Voort, Zell, Hagedorn, Feng, McIntyre, Haghipour, Graf Pannatier and Eglinton2017; Gies et al. Reference Gies, Hagedorn, Lupke, Montluçon, Haghipour, van der Voort and Eglinton2021). A better understanding of what molecules comprise this deep biomass C pool should be explored in future work.
Due to the natural decrease in microbial activity at depth, it can be difficult to produce enough C for a robust AMS measurement using either incubation or extraction methods. Even with a large mass of soil, soil incubations often need to run for months during which time microbial community diversity may shift, creating artifacts and biasing the results, and lengthy experiments can be problematic for some researchers (Schädel et al. Reference Schädel, Beem-Miller, Aziz Rad, Crow, Hicks Pries, Ernakovich, Hoyt, Plante, Stoner, Treat and Sierra2020). For the chloroform biomass extraction method, the issue of low C recovery at depth can be circumvented by extracting from a larger soil mass, thereby increasing the amount of extracted biomass. However, scaling up the extraction also increases the amount of active time required to process the sample. We found that simply doubling the amount of soil and water/chloroform in a single extraction significantly reduced the rate of filtration. Instead, we opted to pool extracts from multiple separate extractions, thereby maintaining a standard time and filter volume for each extraction. While we were able to identify and eliminate some sources of 14C contamination, we were unsuccessful in completely eliminating it. We hypothesize that some contribution to the blank may originate from the polycarbonate filter used to remove fine particles (0.2 μm). Binder-free glass fiber filters at this pore size were not available, however testing without these filters resulted in large amounts of colloidal material passing into the filtrate and skewing the Δ14C values.
CONCLUSIONS
Understanding the role of microbial communities in soil C cycling and the persistence of soil organic matter is challenging given the heterogenous and complex nature of soils. While natural abundance 14C laboratory incubations have some drawbacks, they have provided valuable insight into microbial decomposition and assimilation of soil C. However, additional methods are needed to provide a more direct and mechanistic understanding of microbial C assimilation. The 14C chloroform biomass extraction method we present here can be a useful alternative to soil incubations, possibly avoiding some of the artifacts associated with incubations, though additional research will be needed to assess the inclusion of non-living cells during biomass extraction. Additional methods for isolating specific, short-lived biomolecules, such as RNA, may be required to unambiguously determine the ∆14C of organic molecules being assimilated by active microbial communities.
ACKNOWLEDGMENTS
The authors thank EW Slessarev for conversations and advice regarding the direct chloroform extraction method and L Iancu for independently testing the method. We thank the staff at the Hopland Research and Extension Center who manage the experiment site and Z Kagely for his assistance in digging the soil pit. We acknowledge the traditional, ancestral, unceded territory of the Shóqowa and Hopland People, on which this research was conducted. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 and was supported by the LLNL LDRD Program under Project No. 21-ERD-021 and by the U.S. Department of Energy, Office of Biological and Environmental Research, Genomic Sciences Program LLNL ‘Microbes Persist’ Scientific Focus Area (award #SCW1632). LLNL-JRNL-843130.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit https://doi.org/10.1017/RDC.2023.80
Open access
