Oxygen isotope exchange between water and carbon dioxide in soils is controlled by pH, nitrate availability and microbial biomass through links to carbonic anhydrase activity

is controlled by pH, nitrate availability and microbial biomass through links to carbonic anhydrase activity Sam P. Jones, Aurore Kaisermann, Jerome Ogee, Steven Wohl, Alexander W. Cheesman, Lucas A. Cernusak, and Lisa Wingate 1 INRA, UMR ISPA, 33140, Villenave d’Ornon, France 2 Instituto Nacional de Pesquisas da Amazônia, Manaus – AM, CEP 69060-001, Brasil 3 College of Science and Engineering, James Cook University, Cairns, Queensland, Australia


Introduction
The rate of oxygen isotope exchange (kiso) between soil water and carbon dioxide (CO2) is a key uncertainty in estimating the gross contribution of terrestrial uptake and release to net land-atmosphere carbon exchange from the oxygen isotope composition (δ 18 O) of atmospheric CO2 (Wingate et al., 2009;Welp et al. 2011). The δ 18 O of atmospheric CO2 can be used to trace these large and opposing fluxes because the δ 18 O of leaf-atmosphere CO2 exchange tends to be enriched and distinct compared to well mixed atmospheric CO2 and thus has the potential to serve as an independent tracer of gross primary production (Francey & Tans, 1987). This is the case because the leaves of plants contain considerable concentrations of carbonic anhydrases, which catalyse the hydration of aqueous CO2 and in turn the exchange of oxygen isotopes with water molecules, causing CO2 that interacts with a leaf but is not fixed to inherit the isotopic composition of the leaf water pool (Gillon & Yakir, 2001). As leaf water pools are small and undergo considerable enrichment during evaporation the δ 18 O of this water is enriched relative to that of the soil water and thus CO2 that has interacted with leaves has a distinct isotopic signature in the atmosphere (Francey & Tans, 1987). However, the presence of carbonic anhydrases is not limited to leaves with a number of forms also found in soils (Meredith et al., 2019), but the abundance and activity of these enzymes is poorly known and the degree to which their influence on kiso alters the δ 18 O of atmospheric CO2 is not well constrained. As such, improved understanding of variations in soil kiso benefits efforts to constrain variability and controls on carbon exchange at the ecosystem-level and above.
intra-extra cellular dissolved inorganic carbon gradients (Smith & Ferry, 2000) in response to changes in extra-cellular CO2 and bicarbonate availability (Figure 1 b). Such a control may be supported by the fact both bacteria and fungi grown under CO2 limited conditions have been shown to increase the expression of carbonic anhydrases (Kaur et al., 2009;Kozliak et al., 1995;Merlin et al., 2003). Indeed, such a response in order to maintain the supply of CO2 in bicarbonate dominated systems has been documented for aquatic phototrophs (Hopkinson et al., 2013). The chemistry of non-carbon anions may also play a role in controlling the activity of carbonic anhydrases (Tibell et al., 1984) with the presence of phosphate for example potentially inhibiting extra-cellular carbonic anhydrase activity in soil solutions and thus decreasing k iso (Sauze et al., 2018).
In this respect, the fact that nitrate (NO3 − ) has been shown to inhibit carbonic anhydrases (Peltier et al., 1995) suggests that the inorganic nitrogen chemistry of soil solutions may exert a control, particularly in the context of fertilised agriculture soils or increased atmospheric nitrogen deposition. Indeed, the inhibition of soil carbonic anhyrases by nitrogen fertilisation has been inferred from measurements of carbonyl sulphide exchange (Kaisermann et al., 2018b), but the influence on kiso has yet to be considered.
Here we investigate variations in the rate of oxygen isotope exchange, kiso, through controlled laboratory gas exchange measurements on soil incubations. To understand the drivers of these variations we measured soils, with different chemical and physical properties, sampled from 44 sites across western Eurasia and northeastern Australia. We also conducted a fertilisation experiment on a subset of these soils to investigate the influence of changes in nitrogen availability. We tested three specific, non-exclusive, hypotheses; 1) kiso increases with increases in microbial biomass reflecting the common nature of carbonic anhydrase expression by soil organisms (H1), 2) kiso increases with increases in soil pH reflecting an increase in the amount or efficiency of expressed carbonic anhydrases either because of the response of organisms to unfavourable gradients in intra-extra cellular dissolved inorganic carbon under alkaline conditions or a shift in active functional groups (H2), and 3) kiso will decrease with increases in NO3 − availability as it binds with carbonic anhydrases and directly inhibits enzymatic activity (H3). For the Eurasian soils we also compare these drivers to the predictive power of relatively invariant soil properties that might be used to estimate the rate of exchange in soils at the regional scale and above as required by efforts to better constrain gross primary production.

Methods
To investigate the outlined hypotheses two similar measurement campaigns, each consisting of a spatial survey and an ammonium nitrate (NH4NO3) addition experiment, were conducted aimed at characterising the controls on variations in the rate of oxygen isotope exchange, kiso, across soils from a wide range of environments. In both cases we estimated kiso from gas exchange and soil physical property measurements (Jones et al., 2017;Sauze et al., 2018) and subsequently measured the pH, microbial biomass, NO3 − availability and NH4 + availability of the incubated soils to investigate the controls on this 4 95 100 105 activity. The first campaign focused on soils sampled from across western Eurasia (EUR) whilst the second (AUS) focused on soils sampled in north Queensland, Australia. Sampling sites were broadly classified based on the principle land-cover reported by previous studies or observed during sampling and climatic zone as indicted by the Köppen-Geiger climate classification map of Kottek et al., (2006) and Rubel et al., (2017).

Soil sampling and incubation preparation
For the EUR campaign, collaborators (see acknowledgements) sampled the superficial 10 cm of soil at three locations within 27 sites during the Northern hemisphere summer of 2016 and shipped the soil samples to the Bordeaux-Aquitaine Center of the National Institute of Agricultural Research, France (Figure S1 a). These sites fell within Subarctic (Dfc; n= 6 ), Temperate oceanic (Cfb; n= 13), Hot-summer Mediterranean (Csa; n = 7) and Hot semi-arid (Bsh; n = 1) climate zones and were principally found in forests (n = 16) and grasslands (n = 6). The other remaining sites were located in an agricultural field (n = 1), a peatland (n = 1) and orchards (n = 3). Upon receipt, samples were passed through a 4 mm sieve and mixed to create one homogeneous sample for each site. These soils were stored at 4 °C. A sub-sample of each of these soils was used to determine the initial water content and the soil water holding capacity (Haney & Haney, 2010). For each soil three replicate incubations were prepared with glass jars of 15.54 cm in height and an internal diameter of 8.74 cm. A jar was filled with the wet weight equivalent of 115 to 300 g of dry soil and the water content adjusted to 30 % of the water holding capacity to create a soil column with a surface area of 60.0 cm 2 and a depth of approximately 4 to 7 cm. The jar was then pre-incubated in a climate-controlled cabinet (MD1400, Snijders, Tillburg, NL) for two weeks in the dark at 22 ± 1 °C. This cabinet was continuously flushed with approximately 20 L min −1 of ambient air provided by a pump with an inlet outside of the building to avoid exposing the soil to elevated CO2 concentrations found within the laboratory. During this period, soil water content was periodically adjusted to account for evaporation. Approximately 18 hours prior to measurement the jar was closed with a screw-tight glass lid equipped with inlet and outlet connections and flushed at 250 mL min −1 with dry, synthetic air to promote steady-state conditions. This flow was produced using an in-house dilution system that mixed pure CO2 from a cylinder into CO2-free air generated by an air compressor (FM2 Atlas Copto, Nacka, Sweden) equipped with a scrubbing column (Ecodry K-MT6, Parker Hannifin, USA). This system was set to achieve a CO2 concentration of 400 ± 5 ppm and, reflecting the origin of the CO2 in the cylinder used, had a δ 18 O of approximately −25 ‰ VPDBg. Subsequently the jar was removed to conduct gas exchange and soil property measurements.
For the AUS campaign, we sampled the superficial 10 cm of soil at four locations within 17 sites during July of 2017 and returned these samples to the Cairns campus of James Cook University (Figure S1 b). These sites fell within Tropical monsoon (Am; n = 3), Humid subtropical (Cfa; n = 9) and Monsoon-influenced humid subtropical (Cwa; n= 5) climate zones and were principally found in forests (n = 9) and savannas (n = 6), with the other remaining sites located in a pasture 5 125 130 135 140 145 (n = 1) and a stunted shrub-rich forest (n = 1). These soils were passed through a 4 mm sieve and mixed to create a homogenous sample for each site. A sub-sample of each of these soils was used to determine the initial water content and estimate the re-packed bulk density of the soils. As with the EUR campaign, three replicate incubations were prepared in glass jars for each soil. These jars had a height of 11.56 cm and an internal diameter of 7.45 cm. A jar was filled with the wet weight equivalent of 215 to 450 g of dry soil and the water content adjusted to 30 % water-filled pore space to create a soil column with a surface area of 43.5 cm 2 and a depth of approximately 8.5 cm. The jar was then pre-incubated in an insulated box for one week in the dark at 23 ± 1 °C with periodic adjustments to the water content to account for evaporation. This box was continuously flushed with approximately 10 L min −1 of ambient air provided by an air compressor that serviced building wide laboratory air distribution. The concentration of CO2 in this air was approximately 420 ppm and, and reflecting it's atmospheric origin, had a δ 18 O of approximately 0 ‰ VPDBg. Following pre-incubation the jar was removed to conduct gas exchange and soil property measurements.
An NH4NO3 addition experiment was conducted in both campaigns. To do so an additional three replica incubations were prepared as described above, with these untreated soils serving as controls, for nine and five soils of the EUR and AUS campaigns respectively. Prior to the pre-incubation step, 0.7 mg of NH4NO3 g dry soil −1 was added dissolved in the water used to adjust the water content. This quantity was chosen following Ramirez et al., (2012) to approximate a treatment comparable to typical field studies.

Gas exchange measurements
Gas exchange measurements were made using a similar experimental set-up to that described in Jones et al., (2017). Each jar was connected to a gas delivery system that supplied one of two gas sources, δb,atm or δb,mix, to its inlet. The first inlet condition, δb,atm, consisted of a continuous flow of atmospheric air pumped from an external buffer volume, through a Drierite column (W. A. Hammond DRIERITE Co. LTD, USA) to dry the air and directly to the inlet of the jar. The second condition, δb,mix, was produced by a second continuous flow of atmospheric air pumped from the buffer, through a soda lime column to remove CO2 and a second Drierite column. A mass-flow controller was used to dilute pure CO2 from a cylinder into this dry CO2 free air and then this mix was supplied to the inlet of the jar. The flow rate of pure CO 2 was controlled to match the concentration of the CO2 in δb,mix to that of δb,atm using a control loop feedback based on the difference in concentration between sub-samples of both flows measured with an infra-red CO2 analyser (Li-6262, LI-COR Biosciences, USA). By doing so the principal difference between the two conditions was the isotopic composition of the CO2 present reflecting its origin in the atmosphere (δ 18 O-CO2 of δb,atm = −1.41 ± 2.17 ‰ VPDBg) or a cylinder (δ 18 O-CO2 of δb,mix = −25.33 ± 0.30 ‰ VPDBg). The delivery of either gas to the inlet of the jar was operated by a valve manifold and micro-6 155 160 165 170 controller. Following the manifold, the selected gas stream was split into a chamber line, to which the jar was connected, and a bypass line that terminated at open splits in front of a valve connected to the sample inlet of a CO2 isotope ratio infrared spectrometer (Delta Ray IRIS, Thermo Fischer Scientific, Germany). The flow rate of the chamber line was limited to 171.48 μmol s mol s −1 using a mass-flow controller. The micro-controller was set to supply first one inlet condition through the manifold to the chamber and bypass line and then switch to the second inlet condition. Both inlet conditions were supplied for either 32 (EUR) or 34 (AUS) minutes. The first 20 (EUR) or 22 (AUS) minutes under each condition were used to flush the system and promote steady-state conditions in the incubation jar. After this period, the final 12 minutes during which the condition was supplied before switching was used for gas-exchange measurements. During this 12 minute measurement period the valve in front of the IRIS switched three times between the chamber and bypass line at two minute intervals. Reflecting the pre-incubation conditions, measurements for EUR began with δb,mix as the inlet condition before switching to δb,atm, whilst for AUS the sequence began with δb,atm and then switched to δb,mix. For EUR, the calibration cylinders (21 % O2 and 0.93 % Ar in a N2 balance, Deuste Steinger GmbH, Germany) had total concentration, carbon isotope composition and δ 18 O of CO2, respectively, of 380.26 ppm , −3.06 ‰ VPDB, and −14.63 ‰ VPDBg for the first cylinder, and 481.62 ppm , −3.07 ‰ VPDB and 14.70 ‰ VPDBg for the second cylinder (IsoLab, Max Planck Institute for Biogeochemistry, Germany).
The net CO2 flux, FR (μmol s mol m −2 s −1 ), was calculated from corrected values for the three pairs of chamber and bypass line measurements made at each inlet condition following Eq. (1): where u is the flow rate (mol m −3 s −1 ) through the chamber line, Cc is the total CO2 concentration (ppm) of the chamber line,

Soil properties
After being disconnected from the gas exchange system, a jar was weighed to determine the wet weight of the incubated soil and the total soil depth, zmax (m), measured using a caliper. Soil was then removed from the jar to determine soil water content, pH, microbial biomass, NO3 − availability and NH4 + availability. Soil water contents were determined gravimetrically for sub-samples based on water loss after oven drying for 24 hours at 105 °C. In the EUR campaign, soil water content was determined for three, 1.5 cm thick intervals between 0.0 and 4.5 cm depth. An average gravimetric water content (g g dry soil −1 ) was calculated for the soil column after weighting by total soil depth. In the AUS campaign, soil water content was determined for a single sample covering the total soil depth. Soil bulk density (g cm −3 ) was calculated from the gravimetric water content, the wet weight of the soil in the jar and the volume of the soil column. Total porosity, ϕ t, was calculated from bulk density assuming a particle density of 2.65 g cm −3 (Linn & Doran, 1984). Volumetric water content, θw (m 3 m −3 ), was calculated as the product of gravimetric water content and bulk density. The soil air-filled porosity, ϕ a, was calculated as the difference between the total porosity and volumetric water content. The remaining soil column in the jar was then mixed and sub-samples were taken to determine pH, microbial biomass, NO3 − availability and NH4 + availability. Soil pH was determined in a slurry with a dry weight equivalent soil-to-water ratio of 1:5. Soil microbial biomass (μmol s g C g dry soil −1 ) was determined based on the difference between dissolved carbon extracted from non-fumigated and chloroform-fumigated subsamples using a slurry with a dry weight equivalent soil-to-potassium sulphate solution (0.5 M) ratio of 1:5 and an extraction efficiency value of 0.35. Available NO3 − (μmol s g N g dry soil −1 ) and NH4 + (μmol s g N g dry soil −1 ) were extracted in a slurry with a dry weight equivalent soil-to-potassium chloride solution (1 M where zmax (m) is the total soil-column depth, κ is soil tortuosity calculated here following the formulation of Moldrup et al. (2003) for repacked soils, D (m 2 s −1 ) is the diffusivity of 12 C 16 O 18 O in air (Massman, 1998;Tans, 1998) and ϕa is the air-filled porosity of the soil (see Sauze et al., (2018) for the derivation). Subsequently kiso (s −1 ) was calculated following Eq. (5): where B (m 3 m −3 ) is the Bunsen solubility coefficient for CO2 in water (Weiss, 1974) and θw (m 3 m −3 ) is the soil volumetric water content.

Statistical analyses
Statistical analyses were conducted in R version 3.5 (R Core Team, 2019). Of the 174 individual incubations prepared, 10 were excluded from the dataset because a record for one of the variables of interest; the rate, kiso, of oxygen isotope exchange, pH, microbial biomass, NO3 − availability or NH4 + availability, was missing. For the remaining 164 incubations 9 245 250 255 with complete records, these variables were averaged by sampling site and, for the relevant subset, by whether they received a NH4NO3 addition.
The resultant dataset consisted of mean observations for 44 untreated soils (n = 27 / EUR and 17 / AUS) and 14 soils (n = 9 / EUR and 5 / AUS) that received a NH4NO3 addition. Spatial controls on kiso were investigated across the means of untreated soils. Correlations between kiso, pH, microbial biomass, NO3 − availability and NH4 + availability were investigated through the Spearman's rank correlation between pairs of variables. To test the outlined hypotheses, a multiple generalised linear modelling approach was used to investigate which variables best explained variations in kiso (Thomas et al., 2017). As pH and NH4 + availability were strongly negatively correlated (Spearman's ρ = −0.73), presumably reflecting the pH dependency of NH4 + and ammonia speciation, these were not considered together in the same model whilst all other possible combinations, including sampling campaign (EUR or AUS) to test for the undue influence of systematic experimental differences, were tested. Combinations were limited to models containing four or less predictive terms to prevent over-fitting and each independent variable was centered and scaled to facilitate comparison among the different measurement scales. The model structure and predictive terms included in the minimal adequate model required to explain variations in k iso were selected based on comparison of sample size corrected Aikake's Information Criterion (AICc) and visual assessment of the conformity of model residuals to the assumptions of normality, homogeneity and the absence of unduly influential observations. This model was subsequently re-fitted with the original unstandardised variables. The same approach, limited to two-term models, was also applied to only the 27 soils from the EUR sampling campaign and extended to consider the relationships with soil texture and carbon and nitrogen contents to investigate their utility in upscaling efforts.
To investigate the influence of the NH4NO3 addition on the rate of oxygen isotope exchange, kiso, the variables of interest were expressed as the ratio of the mean of the soils that received an addition and that of their respective untreated counterparts with quotients smaller and greater than one respectively indicating a reduction and increase following addition.
Correlations between these fractional changes for kiso, microbial biomass, pH, NO3 − availability and NH4 + availability were investigated through the Spearman's rank correlation between pairs of variables. The minimal adequate, generalised linear model describing the fractional change in kiso across these soils was investigated by comparing the AICc and visual inspection of the residuals for models that considered each independent variable separately to avoid over-fitting.

Variations among untreated soils
Clear differences in the rate, kiso, of oxygen isotope exchange, pH, microbial biomass, NO3 − availability and NH4 + availability were not apparent as a function of sampling site climatic zone or land-cover ( Figure 2). Estimates of kiso ranged from 0.01 to 0.40 s −1 with the greatest rates occurring in soils sampled from hot-summer Mediterranean (Csa) , hot semi-arid (Bsh) and subtropical (Cfa and Cwa) climates (Figure 2 a). Soil pH ranged from 3.9 to 8.6 and were mostly acidic or neutral with alkaline conditions only found for soils sampled from hot-summer Mediterranean (Csa) and hot semi-arid (Bsh) climates ranged from 2.5 to 64.7 μmol s g N g dry soil −1 with greatest availability found in soils sampled from temperate climates.
Individual relationships between pairs of these variables were investigated through Spearman's rank correlation (Table 1).
Based on AICc and visual inspection of model fit and residuals, the structure of the generalised linear model describing variations in the rate of oxygen isotope exchange, kiso, as the response variable was specified with a gaussian error distribution and log-link function (Thomas et al., 2017). The minimal adequate model with this structure ( Figure S2) included the additive effects of soil pH (0.122), the natural logarithm of NO3 − availability (−0.730) and the natural logarithm of microbial biomass (0.463), the interaction between soil pH and the natural logarithm of NO3 − availability (0.109) and an intercept term (−6.046). This model explained 71 % of the deviance in kiso (Figure 4 a) compared to the null model containing only an intercept term. The best model had an AICc that was 6.1 lower than the next best alternative model which omitted the interaction term, 7.1 lower than the closest model containing sampling campaign and 13.3 lower than the closest model containing the natural logarithm of NH4 + availability. The AICc values of single-term models containing only pH or the natural logarithms of microbial biomass or NO3 − availability were respectively 21.6, 43.6, and 50.2 greater than the best model. The selected model predicts the response variable, kiso-pred (s −1 ), in the original measurement units following Eq. where pH is soil pH, NO3 − is NO3 − availability (μmol s g N g dry soil −1 ) and MB is microbial biomass (μmol s g C g dry soil −1 ). The model predicts that variations in kiso result from positive correlations with soil pH (Figure 3 a) and microbial biomass (Figure   3 c) and negative correlation with NO3 − availability. The interaction between soil pH and NO3 − availability is such that the negative influence of NO3 − on kiso occurs mainly under acidic conditions and is marginal at neutral to alkaline pH (Figure 3 b).
As with the full dataset, across the 27 soils from the EUR sampling campaign the strongest relationship with the rate of oxygen isotope exchange, kiso, was found with pH (Spearman's ρ = 0.58), whilst a weaker but still significant (

Variations induced by NH4NO3 addition
The addition of NH4NO3 systematically increased available NO3 − and NH4 + and decreased the rate of oxygen isotope exchange, kiso, and soil pH. Available NO3 − and NH4 + in the treated soils that received the NH4NO3 addition were respectively 1.9 to 173.6 and 3.7 to 18.8 times greater than in the corresponding untreated soils. Soil pH and k iso were respectively 0.86 to 0.98 and 0.21 to 0.76 times smaller in the soils that received the addition than in the corresponding untreated soils. The addition did not have a systematic influence on microbial biomass, which varied between 0.64 and 1.84 of the magnitude in the corresponding untreated soils.
Individual relationships between pairs of these fractional changes were investigated through Spearman's rank correlation (Table 2). Strong, significant correlations (p < 0.05) for variable pairs were found between the fractional changes in kiso and Based on AICc and visual inspection of model fit and residuals, the structure of the generalised linear model describing variations in the fractional change in the rate of oxygen isotope exchange, kiso, as the response variable was specified with a betareg error distribution and identity link function (Thomas et al., 2017). The minimal adequate, single term model with this structure included the natural logarithm of the fractional change in NO3 − availability (−0.499) and an intercept term (1.219).
This model predicts the variations in the fractional change in kiso following NH4NO3 addition across soils from the 14 sites considered result from a negative relationship with fractional changes in NO3 − availability ( Figure 5). This relationship explained 76 % of the deviance in the fractional change in kiso and the model had an AICc that was 13.2 lower than the next best alternative model which included the fractional change in soil pH and an intercept term.

Discussion
This study aimed to investigate the drivers of variations in the rate of oxygen isotope exchange, kiso, between soil water and CO2 with a view to improving our ability to predict the influence of soils on the δ 18 O of atmospheric CO2 and our understanding of dynamics in the activity of carbonic anhydrases expressed by soil microbial communities. To do so, controlled incubation experiments were conducted with soils sampled from 44 sites across western Eurasia and northeastern Australia in order to estimate kiso and metrics relating to hypothesised controls on this activity. Estimates of kiso for untreated soils ranged from 0.01 to 0.4 s −1 (Figure 2 a). In all cases these rates exceeded theoretical uncatalysed rates of oxygen isotope exchange calculated for the incubation conditions (Uchikawa & Zeebe, 2012), which ranged from 0.00008 to 0.008 s −1 , indicating the presence of active carbonic anhydrases. These observations, with a median of 0.07 s −1 , are in good agreement with a number of previous studies which estimated kiso ranging from 0.03 to 0.15 s −1 for sieved soils incubated in the dark (Jones et al., 2017;Sauze et al., 2018Sauze et al., , 2017, but are somewhat lower than those reported by Meredith et al. (2019) with a median and range of 0.46 s −1 and 0.08 to 0.88 s −1 , respectively. These greater kiso are more comparable to those, ranging from 0.01 to 0.75 s −1 , reported by Sauze et al. (2017) for soils with well developed algal communities. Direct comparison with field observations is non-trivial because these older studies tend to address soil carbonic anhydrase activity as a range of enhancement factors over a temperature sensitive uncatalysed rate of hydration (Seibt et al., 2006;Wingate et al., 2008Wingate et al., , 2009Wingate et al., , 2010. However, using the mid-point of the enhancement factors and soil temperatures reported by Wingate et al. (2009), we can estimate that kiso varied between 0.04 and 13 s −1 with a median of 0.31 s −1 across the seven ecosystems studied. Whether the potential for kiso to be orders of magnitude greater in the field than in incubation studies is an artefact of the sensitivity of the methodology applied to estimate the isotopic composition of the soil water pool from which exchanged We hypothesised that the rate of oxygen isotope exchange, kiso, might be positively correlated with microbial biomass (H1), positively correlated with soil pH (H2) and negatively correlated with NO3 − availability (H3). We found evidence in support of all three hypotheses with the minimal adequate statistical model explaining variations in kiso observed across untreated soils including all three of these terms (Eq. 6). The model suggests that the positive relationship with soil pH (Figure 3 a), the strongest single predictor of variations in kiso, reinforces the emergent view of soil pH as the principal driver of variations in carbonic anhydrase expression by soil microbial communities (Sauze et al., 2018). Marked increases in kiso under alkaline conditions likely reflects a shift in microbial community towards organisms that express more or more efficient carbonic anhydrases than those found under acidic conditions (Meredith et al., 2019;Sauze et al., 2018Sauze et al., , 2017 and the need for organisms to up-regulate carbonic anhydrases expression. This may be required in order to control the transport and availability of CO2 and bicarbonate in response to the pH dependent speciation of dissolved inorganic carbon (Figure 1 b) as has been observed for both intra-and extra-cellular carbonic anhydrase activity in non-soil settings (Hopkinson et al., 2013;Kaur et al., 2009;Kozliak et al., 1995;Merlin et al., 2003). Similarly, in the positive relationship with microbial biomass (Figure 3 c) we find support for a secondary role for the expected link between the abundance of organisms likely to be expressing carbonic anhydrase and kiso for a given set of biogeochemical conditions (Sauze et al., 2017). Finally, through the negative relationship with NO3 − availability (Figure 3 b) we show for the first time that kiso in soils is sensitive to dissolved inorganic nitrogen chemistry. Outwith soils, anions including NO3 − have been shown to inhibit carbonic anhydrase activity by binding with the enzyme (Peltier et al., 1995;Tibell et al., 1984). The fact that this binding and subsequent inhibition of carbonic anhydrase activity has been shown to be more efficient under acidic conditions but have minimal influence at high pH may reflect the role of protonation in this behaviour (Johansson & Forsman, 1993, 1994. Interestingly, the interaction between soil pH and NO3 − availability identified here, leading to a larger negative influence of NO3 − availability under acidic conditions (Figure 3 b), is in agreement with this observation. This suggests that the influence of NO 3 − availability on carbonic anhydrases activity is likely minimal in neutral and alkaline soils and the constraints imposed by pH and microbial community size are of greater importance. To better understand the relationship between k iso and soil inorganic nitrogen we conducted an NH4NO3 addition experiment. As in other studies, the addition of NH4NO3 not only increased the availability of NO3 − and NH4 + but also acted to decrease soil pH and caused non-systematic changes in microbial biomass . Reflecting the different magnitudes of these changes, the observed decrease in kiso in soils receiving the addition relative to their untreated counterparts was best explained by the increase in NO3 − availability ( Figure 5). Notably the weak relationship between changes in kiso and NH4 + availability identified in this experiment (Table 2) suggests the relationship between these variables across the untreated soils (Table 1) does indeed reflect the pH sensitivity of ammonia speciation rather than a direct causal link. The negative relationship between NO3 − availability and kiso appears to support the proposed mechanism of carbonic anhydrases inhibition. However, an alternative explanation, invoked to explain reductions in the activity of enzymes involved in nitrogen acquisition following fertilisation , may be that carbonic 14 385 390 395 400 405 anhydrases play some role in the soil nitrogen cycle that is alleviated by increases in NO3 − availability following NH4NO3 addition and thus leads to down-regulation of expression (DiMario et al., 2017;Kalloniati et al., 2009;Rigobello Masini et -Masini al., 2006). Indeed, such a function would help explain why the microbial communities in the untreated acidic, higher relative to lower NO3 − availability soils do not appear to need to compensate for the inhibition of carbonic anhydrases as we might expect from the economic theory of enzyme investment if they are facilitating important metabolic reactions (Burns et al., 2013). Much needed development of our understanding of the intra-and extra-cellular distribution of soil carbonic anhydrases and their relationship to spatial and temporal variations in chemical conditions experienced by the microbial communities that express them are required to confirm the mechanistic link among these observations. Improvements to our ability to predict the influence of soils on the the δ 18 O of atmospheric CO2 are important in refining the use of this tracer to constrain gross primary production at the ecosystem-scale and above (Wingate et al., 2009;Welp et al., 2011) . The absence of strong patterns with climate or land-cover in this study may well reflect the fact that the temperature and moisture conditions used are unrepresentative of field conditions especially for colder and drier sites (Figure 2 a).
Whether or not up-scaling based on such classes is feasible is somewhat unknown (Wingate et al., 2009). However, the data reported here does provide the basis for an empirical approach to predicting the rate of oxygen isotope exchange, kiso, for a given soil (Figure 3). The minimal adequate statistical model described (Eq. 6) was able to provide broadly unbiased predictions of variations observed in kiso across the untreated soils of the 44 sites considered (Figure 4 a). Indeed, broad agreement between predictions of the fractional changes in kiso between untreated and treated, which were not used in model selection, soils following the NH4NO3 addition encouragingly suggest that this model could be used to provide reasonable predictions of kiso for other soils (Figure 4 b). More observations from alkaline soils are required to reduce uncertainty found at greater kiso and further validation is required to avoid biased predictions outside of the ranges considered ( Figure 3). A significant challenge to using this relationship to predict kiso is likely the availability of suitable pedotransfer functions, particularly for NO3 − availability and microbial biomass, to estimate patterns in the proposed drivers (Van Looy et al., 2017).
Given the interaction between soil pH and NO3 − availability (Figure 3 a & b), the absence of such data may not seriously compromise predictions for fertilised agricultural soils which are typically not strongly acidic. However, accurately predicting natural spatial and seasonal variability and the influence of future changes in atmospheric NO3 − deposition (DeForest et al., 2004) may be more problematic. For this reason we considered whether more readily available parameters such as soil texture, carbon content and nitrogen content might provide an alternative basis for empirical predictions of kiso (Van Looy et al., 2017). Relationships between these variables and kiso were relatively weak and could only explain a marginal amount of the observed variability. Considering these properties in combination with soil pH yielded clay content as a secondary significant term potentially reflecting a relatively strong co-correlation (Spearman's ρ = 0.5) with NO3 − availability. Soil pH and clay content may provide an alternative empirical approach to predicting kiso when the availability of soil property data is limited.