Quantifying microbial growth and carbon use efficiency in dry soil environments via 18O water vapor equilibration

Abstract Soil microbial physiology controls large fluxes of C to the atmosphere, thus, improving our ability to accurately quantify microbial physiology in soil is essential. However, current methods to determine microbial C metabolism require liquid water addition, which makes it practically impossible to measure microbial physiology in dry soil samples without stimulating microbial growth and respiration (namely, the “Birch effect”). We developed a new method based on in vivo 18O‐water vapor equilibration to minimize soil rewetting effects. This method allows the isotopic labeling of soil water without direct liquid water addition. This was compared to the main current method (direct 18O‐liquid water addition) in moist and air‐dry soils. We determined the time kinetics and calculated the average 18O enrichment of soil water over incubation time, which is necessary to calculate microbial growth from 18O incorporation in genomic DNA. We tested isotopic equilibration patterns in three natural and six artificially constructed soils covering a wide range of soil texture and soil organic matter content. We then measured microbial growth, respiration and carbon use efficiency (CUE) in three natural soils (either air‐dry or moist). The proposed 18O‐vapor equilibration method provided similar results as the current method of liquid 18O‐water addition when used for moist soils. However, when applied to air‐dry soils the liquid 18O‐water addition method overestimated growth by up to 250%, respiration by up to 500%, and underestimated CUE by up to 40%. We finally describe the new insights into biogeochemical cycling of C that the new method can help uncover, and we consider a range of questions regarding microbial physiology and its response to global change that can now be addressed.

. Microbial respiration, growth, turnover, and carbon use efficiency (CUE), that is, the proportion of C taken up by microorganisms that is allocated to growth, are key parameters of soil microbial C metabolism. Microbial CUE can be indicative of soil C sequestration (Bradford & Crowther, 2013). Accurate quantification of these parameters and their response to changing environmental conditions is essential for parameterization of models to predict future soil C stocks as well as to develop management practices to promote soil C sequestration (Kallenbach, Wallenstein, Schipanksi, & Grandy, 2019;Li et al., 2019). Under drought, microbial physiology can be altered as soil microorganisms become disconnected from their substrates (de Nijs, Hicks, Leizeaga, Tietema, & Rousk, 2019;Moyano, Manzoni, & Chenu, 2013;Schimel, Balser, & Wallenstein, 2007). Microorganisms must be in contact with soil water to remain active. Because of their semipermeable cell membrane, some microorganisms need to produce osmolytes to reduce their internal water potential and to avoid dehydration and death when soil moisture is low (Borken & Matzner, 2009;Schimel, 2018). Despite these restrictions, soil microorganisms can maintain high levels of activity under drought, as shown by soil respiration and transcriptomics analyses (Roy Chowdhury et al., 2019;Schimel, 2018). Soil microorganisms can maintain activity at much lower water potentials (even lower than −15 MPa) than plants do, as microbial cells interact with smaller soil pores (10-100 µm in size) that might retain hydraulic connectivity under dry conditions, despite negligible diffusivity at the macroscale (Manzoni & Katul, 2014). The different sensitivity of plant CO 2 assimilation and ecosystem respiration under drought (Schwalm et al., 2010) can cause an ecosystem to turn from a carbon sink into a carbon source (Hoover & Rogers, 2016;Jarvis et al., 2007;Schimel, 2018). While plant responses to drought are relatively well understood, physiological responses of soil microbial communities to drought have remained largely elusive.
The most reliable tools to quantify microbial community-level physiological processes in soils are based on stable isotope approaches (Dumont & Murrell, 2005). However, classical stable isotope techniques developed for liquid samples or pure cultures fail due to the complexity of the soil matrix. Recently developed methods can quantify essential parameters of microbial physiology, such as microbial growth and CUE (Blazewicz & Schwartz, 2011;Brant, Sulzman, & Myrold, 2006;Spohn, Klaus, Wanek, & Richter, 2016;Zheng et al., 2019). The common ground of these current methods is the addition of 13 C or 18 O tracers in liquid form (Geyer, Dijkstra, Sinsabaugh, & Frey, 2019). The addition of 18 O enriched water, for example, has allowed the substrate independent quantification of microbial growth and CUE (Figure 1), by tracing 18 O incorporation into genomic DNA (Spohn et al., 2016). However, water added to a dry soil sample causes a burst in microbial activity, as indicated by increases in respiration (up to 500% compared to continuously wet soils; Figure 1). This so called "Birch-effect" (Birch, 1958) can start minutes after rewetting and last for up to 6 days (Canarini, Kiaer, & Dijkstra, 2017;Fraser et al., 2016). The source of this respired C continues to be debated and has been attributed to microbial material (osmolytes or lysed cells), or to mobilization of dissolved organic carbon by aggregate disruption (Canarini et al., 2017;Fraser et al., 2016;Schimel, 2018;Warren, 2014). Similar to respiration, microbial growth is stimulated by re-wetting after drought, although usually delayed by hours relative to the respiratory response (Blazewicz, Schwartz, & Firestone, 2014;de Nijs et al., 2019;Meisner, Bååth, & Rousk, 2013).
Because of the Birch effect, any method that utilizes water to introduce isotope tracers to study microbial physiology is only able to capture the response of microbial respiration, growth, and CUE to rewetting but not to continuously dry conditions. To our knowledge, only one other study has attempted to develop a method to avoid rewetting effects by applying 13 C acetic acid vapor (Herron, Stark, Holt, Hooker, & Cardon, 2009). However, addition of 13 C labeled acetic acid introduces a labile C source that might itself affect microbial growth and respiration (Geyer et al., 2019), and at the same time acetic acid might acidify soils. Here we assessed the validity of a substrate independent method to measure microbial growth in dry soils without changing the soil water content. The method is based on the incorporation of 18 O into soil water by liquid water-vapor isotopic equilibration. Because the isotopic composition of two neighboring liquid water pools in a closed environment will approach an average concentration over time (Urey, 1947), 18 O will equilibrate between the liquid tracer outside the soil and the soil water through evaporation and condensation processes. We assessed the speed of this process and demonstrated the reliability of the method using three different soils by comparison to the currently used methodology (direct 18 O-liquid water addition).  Table S1.

| Experimental setup
Before the experiment, one part of each soil was air-dried to a water content of around 5% (dry mass basis). The air-dried soils as well as the moist soils were kept at room temperature to acclimate for 5 days. After that, we carried out three different tests

| Temporal dynamics of 18 O equilibration
To test whether soil water can be sufficiently enriched with 18 O through water vapor equilibration to allow microbial growth determination in a relatively short time period (24 hr), we set up the fol-

| Simplified method to indirectly quantify equilibration of 18 O in soil water
In order to calculate the time kinetics of isotope equilibration in soil water without having to extract the soil water (see Section 2.3), we determined the precision of an indirect measurement. This was achieved with the same experimental conditions as in Section 2.2.1, but only on air-dry soil samples (as no differences were found in time kinetics between dry and moist conditions) and only for the 18 O-vapor equilibration method (n = 45). At four time points (2,4,8 and 24 hr), the water left at the bottom of the headspace vials was collected and used to analyze its 18 O enrichment (as in Section 2.3).

| Comparison of microbial growth, respiration, and CUE between the two methods
We set up another set of samples to determine microbial growth, respiration and CUE. As in the previous tests, aliquots (400 mg) of sieved soil were weighed in 1.2 ml plastic vials and inserted in 27 ml glass headspace vials. All samples were incubated for 24 hr. Samples included the following: three soils, two soil moisture levels (moist and air-dry), two approaches of tracer addition ( 18 O-vapor equilibration vs. direct 18 O-liquid water addition), and 18 O labeled versus natural isotope abundance samples in three replicates. To the latter, the same volume of non-labeled high purity water was added. For the direct 18 O-liquid water addition method, the dry samples were also subject to two water additions-either the soil was brought to 60% WHC with 18 O labeled water (70-160 µl) or we added only 30 μl of 18 O labeled water (reaching 19%-28% WHC, with a final total enrichment of 20 atom% 18 O in both cases). The low water addition was done to reduce the effects of rewetting as different moisture conditions between the drying phase and the rewetting phase can influence microbial growth and activity (Canarini et al., 2017;Meisner, Leizeaga, Rousk, & Bååth, 2017). This led to a total of 81 samples. We also incubated three replicates of each soil with no water addition, from which only respiration measurements were taken to assess effects of soil vapor absorption on soil respiration.
After their respective incubation times, one gas sample was collected from each headspace vial to measure CO 2 accumulation. Then the headspace vials were opened, the plastic vials containing the soil aliquots collected, closed, shock frozen in liquid nitrogen, and stored at −80 °C until further analyses.

| Extraction of soil water and determination of its 18 O enrichment
To determine the 18 O enrichment in soil water, frozen soil samples from Section 2.2.1 were subjected to cryodistillation, as described in (Plavcova et al., 2018). Briefly, frozen soil samples were transferred to 12 ml glass vials and inserted into a heating block. These were airtightly connected to 300 µl plastic vials sitting upside down in a metal block cooled by liquid N. The heating block was heated to 90 °C and the evaporating water was condensed and frozen in the cooled 300 µl plastic vials. To account for potential isotope fractionation during the extraction, water of five different known 18 O concentrations was treated analogous to the soil samples. Water collected by cryodistillation and from Section 2.2.2 was then analyzed through equilibration of 18 O in H 2 O with CO 2 by a Gasbench II headspace sampler connected to a Delta V Advantage isotope ratio mass spectrometer (Thermo Fisher). where C Growth is the flux of C allocated to biomass production (growth), and C Respiration is the flux of C allocated to the production of CO 2 (respiration). Microbial CUE was then calculated by the following equa-

| Statistical analyses
Statistical differences in microbial respiration, growth, and CUE between tracer addition approaches and soil types were assessed by two-way ANOVA and one-way ANOVA followed by Tukey HSD post hoc tests. When results were not normally distributed or homoscedastic, data were log or rank transformed using the ARTool package (Wobbrock, Findlater, Gergle, & Higgins, 2011 (3) C Uptake = C Growth + C Respiration ,

| Water equilibration dynamics and adaptation of the 18 O-vapor equilibration method
In a closed system of two liquid water sources, the oxygen isotopes will redistribute via the vapor phase to a common concentration.
Here we tested whether the 18 O-vapor equilibration method allows isotopic enrichment of soil water to a similar extent as direct 18 O-liquid water addition (20 at% 18 O) within a relatively short time period (24 hr). Figure 2 ( In an ideal situation, the isotopic exchange between two water pools, that is, an internal soil water pool and an external 18 O water pool, through the vapor phase is controlled by the relative pool sizes of the water pools, the isotopic composition of the water pools involved, and the relative surface areas of these water pools (Ingraham & Criss, 1993). However, soil properties may also affect isotopic exchange rates. For instance, larger aggregate sizes and/or a higher soil porosity may lead to higher water vapor diffusion rates into soils (Jabro, 2009 Table S2) also when used for moist soil conditions ( Figure S4). This allowed us to adopt a rapid and simple method to indirectly calculate the time-averaged 18 O enrichment of soil water that soil microorganisms are exposed to during the experiment. These values need to be measured for each soil type before an experiment, and also if headspace vials with different volume are used, as this might affect the 18 O equilibration time kinetics.

| Methodological differences in microbial growth, respiration, and CUE estimates
By accounting for the time kinetics of 18 O equilibration between external labeled water and soil water (as explained in Section 3.1) we calculated microbial growth in the 18 O-vapor equilibration method. Microbial growth, respiration, and CUE could thus be compared between F I G U R E 2 18 O isotope exchange kinetics of soil water following vapor equilibration (expressed as atom% 18 O) in the three different soils analyzed (from left to right: grassland, forest, and agricultural soil) with incubation time (x-axis). In the upper panels points represent data obtained from cryodistillation extraction of soil water for two soil moisture conditions (air-dry and moist soils) and lines represent the best model fit from Equation (1). The bottom panels represent the model prediction generated from measurements of 18 O kinetics in the labelled external water pool during isotope exchange with soil water via vapor exchange as described in Sections 2.2.2 and 2.4 of Materials and Methods for dry soils only.
18 O-vapor equilibration and direct 18 O-liquid water addition. As expected, dry soils to which water was added directly showed a strong increase in microbial respiration. Dry soils subjected to the 18 O-vapor equilibration method respired three to six times less than rewetted soils (Figure 4a), which had values 346%, 229%, and 516% higher in the grassland, forest, and agricultural site, respectively. We also measured respiration without any addition of water and found no significant difference in respiration rates between untreated soils and soils F I G U R E 3 (a) Respiration rates of grassland, forest, and agricultural soils treated with the 18 O-vapor equilibration method (orange boxes) compared to untreated soils with no direct or indirect water application (white boxes). No significant differences were found following a two-way ANOVA test (F = 1.461; p = .25). Values were log-transformed to meet the assumption of homogeneity of variances. (b) Soil vapor absorption (net weight gain) or soil evaporation (net weight loss) in the three tested soils (grassland, forest, and agricultural soil) at the different soil moisture levels (air-dry, moist) used during the experiment (calculations are described in the Supplementary Methods section).
Negative values indicate no net vapor absorption but rather soil evaporative water loss from moist soils while dry soils gained weight by vapor absorption. This net gain ranged between 0% in dry forest soils, 0.1% in dry grassland soils, and 1.1% in dry agricultural soils (all percentages given relative to initial soil water content). Net gains through soil vapor absorption were therefore negligible.

F I G U R E 4
Soil microbial respiration (a), growth (b), and carbon use efficiency (CUE; c) of a grassland, a forest, and an agricultural site measured in air-dry and moist soils using direct 18 O-liquid water addition at high water addition rate (increasing soil water content to 60% water holding capacity; grey boxes) or with minimal liquid 18 O-water addition (white boxes; dry soils only), and using 18 O-vapor equilibration (orange boxes). Letters indicate statistically significant differences. See Table S3 for full statistical results.
The increase in respiration rates following water addition to dry soils is a common phenomenon (Birch, 1958). While the exact nature of the C released through respiration upon rewetting is still subject to debate, it is known to derive from a combination of abiotic processes, internal use of metabolites, and increased reconnection between available substrates and decomposers (Schimel, 2018;Schimel et al., 2007).
Interestingly, also moist soil obtained from the forest and agricultural sites respired 30%-40% more when water was directly added compared to the 18 O-vapor equilibration method, but not in the grassland soil. Because different soils have different soil moisture ranges at which microorganisms reach optimum activity (Moyano et al., 2012), this result might indicate that adding water causes increases in microbial activity only when the initial soil water content is lower than optimal. Similar to respiration, rewetting of dry soils stimulated microbial growth (Figure 4b). Microbial growth was always higher when water was added directly compared to the 18 O-vapor equilibration method.
The differences were significant in all soils, all water regimes, and for all water amounts added, with the exception of dry forest soils. Here BOX Potential applications of the 18 O-vapor equilibration method to investigate microbial community growth and carbon use efficiency (CUE) in a range of different ecosystems and experimental approaches. only the larger water addition (increasing soil water content to 60% WHC) showed significantly higher growth rates than the 18 O-vapor equilibration method (for full statistical results see Table S3). We could thus show that adding water to dry soils can cause misleading results if used to estimate microbial growth, as microbial growth is known to increase following rewetting of dry soils for both, bacteria and fungi (Hicks, Ang, Leizeaga, & Rousk, 2019). This increase in growth rates following rewetting had values 279%, 92%, and 226% higher in the grassland, forest, and agricultural soil, respectively.
We found no difference in microbial CUE between both methods in moist soils of all sites (Figure 4c). This shows that although growth and respiration may be stimulated by liquid water addition in fresh soils, their increase was proportional and therefore did not affect CUE estimates in moist soils. On the other hand, when liquid water was directly added to dry soils microbial CUE was significantly underestimated compared to the 18 O-vapor equilibration method in forest and agricultural soils (32% and 38% lower, respectively). This indicates that when water is added directly to dry soils, the response of microbial growth and respiration was not proportional. In dry soils the stronger stimulation of respiration than growth by water addition underestimated microbial CUE. Moreover, we found that microbial CUE was significantly higher in dry compared to moist soils (measured by 18 O-vapor equilibration method), indicating a greater sensitivity of respiration than growth to soil drying.

| CON CLUS IONS
Microbial physiology controls large fluxes of C from soil to the atmosphere but also the proportion of C remaining in the soil that can potentially be stabilized. Microbial growth, respiration, and CUE thus require precise quantification to improve predictions of soil C cycling.
A caveat of current approaches to measure microbial physiology is that a tracer is introduced with an aqueous solution, inevitably causing rewetting of dry soils. Here we present a new approach ( 18 O-vapor equilibration) that resolves this issue and expands the possibilities of future studies to accurately quantify microbial growth and CUE in dry soils. The proposed method uses isotopic equilibration between an external 18 O labeled water pool and soil water via the vapor phase and provides similar microbial CUE results as the direct 18 O liquid water addition method when used at near-optimal soil water content.
However, when applied to dry soil the liquid water addition overestimated microbial growth by up to 250%, respiration by up to 500%, and underestimated CUE by up to 40%. The 18 O-vapor equilibration method thus greatly reduces rewetting biases. We further describe new insights into the biogeochemical C cycle that the new method can help uncover (Box 1) and consider a wide range of questions regarding microbial physiology and its response to global change that can now be proposed and addressed.

ACK N OWLED G EM ENTS
The study was funded by the Austrian Science Fund (FWF; project P 30428-B32). Access to the cryodistillation line was supported by the Czech Research Infrastructure for Systems Biology C4SYS (Project no. LM2015055).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.

O RCI D S U PP O RTI N G I N FO R M ATI O N
Additional supporting information may be found online in the