Isotopic Overprinting of Nitrification on Denitrification As a Ubiquitous And

Home , Denitrification

Isotopic overprinting of nitrification on denitrification PNAS PLUS as a ubiquitous and unifying feature of environmental nitrogen cycling

Julie Grangera,1 and Scott D. Wankelb,1

aDepartment of Marine Sciences, University of Connecticut, Groton, CT 06340; and bDepartment of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods Hole, MA 02540

Edited by Donald E. Canfield, Institute of Biology and Nordic Center for Earth Evolution (NordCEE), University of Southern Denmark, Odense M., Denmark, and approved August 26, 2016 (received for review January 25, 2016)

15 Natural abundance nitrogen and oxygen isotopes of nitrate (δ NNO3 invaluable tool to differentiate sources, track their distribution, 18 and δ ONO3) provide an important tool for evaluating sources and and determine the biogeochemical transformations acting on − − transformations of natural and contaminant nitrate (NO3 )inthe NO3 . By convention, isotope ratios are reported using δ notation, − 15 15 14 15 14 18 environment. Nevertheless, conventional interpretations of NO3 where δ N = ([ N/ N]sample/[ N/ N]air − 1) × 1,000 and δ O = 18 16 18 16 isotope distributions appear at odds with patterns emerging from ([ O/ O]sample/[ O/ O]VSMOW − 1) × 1,000, in units of per mille studies of nitrifying and denitrifying bacterial cultures. To resolve (‰). Given two isotopic tracers for a single compound, this this conundrum, we present results from a numerical model of approach can be powerful, as each isotope system provides com- − δ18 NO3 isotope dynamics, demonstrating that deviations in ONO3 plementary information on sources and biogeochemical transfor- 15 vs. δ NNO3 from a trajectory of 1 expected for denitrification are − mations (2). Accurate interpretation of isotope distributions, explained by isotopic over-printing from coincident NO3 production however, strongly hinges on knowledge of the isotope com- by nitrification and/or anammox. The analysis highlights two driving position of source terms and on a rigorous understanding of parameters: (i)theδ18Oofambientwaterand(ii) the relative flux − isotopic discrimination associated with biological transforma- of NO3 production under net denitrifying conditions, whether tions of N pools. The isotopic discrimination associated with SCIENCES catalyzed aerobically or anaerobically. In agreement with existing specific N transformations is quantified by the isotope effect, e, ENVIRONMENTAL > analyses, dual isotopic trajectories 1, characteristic of marine where e (‰) = [(lightk/heavyk) − 1] × 1,000, and k refers to the re- denitrifying systems, arise predominantly under elevated rates of − − > spective specific reaction rate constants of light and heavy iso- NO2 reoxidation relative to NO3 reduction ( 50%) and in associa- topologues (3). Although many of the important source terms and tion with the elevated δ18O of seawater. This result specifically im- − isotope effects of the N cycle are constrained, some remain equiv- plicates aerobic nitrification as the dominant NO producing term in 3 ocal. In particular, recent observations emerging from bacterial and marine denitrifying systems, as stoichiometric constraints indicate − archaeal cultures and from incubations of environmental samples anammox-based NO production cannot account for trajectories − 3 have uncovered isotopic discrimination trends for NO isotopes >1. In contrast, trajectories <1 comprise the majority of model solu- 3 that appear at odds with trends typically ascribed to analogous tions, with those representative of aquifer conditions requiring − biological transformations in soils and aquifers (4–14). This de- lower NO2 reoxidation fluxes (<15%) and the influence of the lower δ18O of freshwater. Accordingly, we suggest that widely ob- velopment has led to conflicting environmental interpretations, 18 15 reflecting a lack of consensus on fundamental isotope systematics of served δ ONO3 vs. δ NNO3 trends in freshwater systems (<1) must − the processes driving the N cycle. Importantly, the discrepancies result from concurrent NO3 production by anammox in anoxic aquifers, a process that has been largely overlooked. Significance nitrate | nitrification | denitrification | isotopes | anammox Stable isotopes of nitrate have long provided a tool for track- he advent of the Haber–Bosch process late in the 19th cen- ing environmental sources and biological transformations. Ttury initiated an unprecedented increase in anthropogenic However, divergent interpretations of fundamental nitrate loading of reactive nitrogen (N) to the biosphere, setting into isotope systematics exist among disciplinary divisions. In an motion cascades of environmental impacts, including eutrophica- effort to transcend disciplinary boundaries of terrestrial and tion and hypoxia, ecosystem acidification, and loss of biodiversity marine biogeochemistry, we use a quantitative model for (1). This intensification of environmental N release from agricul- coupled nitrogen and oxygen isotopes of nitrate founded on tural and industrial activities, power generation, municipal and benchmarks established from microbial cultures, to reconcile septic wastewater, and domestic fertilizer has tremendously al- decades of nitrate isotopic measurements in freshwater and tered the global N cycle, effectively doubling annual global N seawater and move toward a unified understanding of cycling turnover (1). In groundwater, the most common nitrogenous processes and isotope systematics. Our findings indicate that − denitrification operates within the pervasive context of nitrite contaminant is nitrate (NO3 ), with recognized and long-term effects on both human and ecological health. Thus, control and reoxidation mechanisms, specifically highlighting the relative − importance of nitrification in marine denitrifying systems and elimination of NO3 contamination are priorities of environmental and health agencies worldwide. Despite its significance to anammox in groundwater aquifers. − global health and ecosystem function, identifying sources of NO3 , tracing its dispersal and attenuation, and gauging its ecological Author contributions: J.G. and S.D.W. designed research, performed research, analyzed data, and wrote the paper. impact remain challenging. − The authors declare no conflict of interest. Mitigation of NO3 pollution has necessitated identification of its sources and hydrologic flow paths to monitor the fate and This article is a PNAS Direct Submission. 1To whom correspondence may be addressed. Email: [email protected] or sdwankel@ natural attenuation processes occurring in pollutant plumes. To whoi.edu. this end, the natural abundance stable isotope ratios of nitro- This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 15 14 18 16 − gen ( N/ N) and oxygen ( O/ O) in NO3 have provided an 1073/pnas.1601383113/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1601383113 PNAS | Published online October 4, 2016 | E6391–E6400 Downloaded by guest on September 26, 2021 between isotopic trends in environmental systems and those from freshwater environments (39, 40), although this tenet has not been culture-based observations raise the possibility that biogeochemical examined specifically. The potential for analogous biogeochemical − N dynamics inferred from environmental NO3 isotopic measure- dynamics to affect isotope distributions in freshwater and marine ments reflect more complexity than previously realized. systems clearly merits exploring. − − Conventional interpretation schemes for NO3 isotopes differ To arrive at a shared understanding of environmental NO3 from culture observations with regard to the isotope systematic of isotope systematics, we present the results of a multiprocess nu- − − − denitrification, the stepwise reduction of NO3 to NO2 (nitrite), merical model of dual NO3 isotope dynamics parameterized on nitric oxide, nitrous oxide, and finally N2 by heterotrophic bacteria, the basis of fundamental features revealed from culture studies. which is the dominant loss term of reactive N from the biosphere. From this improved understanding of isotopic fractionation during − − Early studies of NO3 isotope dynamics in groundwater docu- redox cycling of N, we explore implications for NO3 production – 15 18 − − mented parallel enrichment of δ Nandδ OofNO3 in associa- by NO2 oxidizing bacteria and by anammox—occurring concur- − − tion with NO3 attenuation from denitrification, approximating a rently with denitrification, specifically focusing on resulting NO3 N linear trajectory with a slope 0.5–0.8 (15–17). Indeed, this salient and O isotope trajectories. We use this framework to evaluate the − trend has long been considered a unique diagnostic signal of de- potential extent of processes other than unidirectional NO3 con- nitrification (2). However, following the advent of the denitrifier sumption by denitrification, which may harbor the key for resolving − method (18, 19), measurements in cultures of both freshwater and the discrepancy between decades of groundwater NO3 observa- marine denitrifying bacteria revealed dual isotope enrichments as- tions and our physiological understanding of the isotope systematics − sociated with assimilatory and dissimilatory NO3 consumption of microbial N cycling. The scenarios explored herein call attention systematically following linear trajectories of ∼1 (9, 10, 12, 19, 20), to the potential influence of N cycling dynamics that have been contrasting with the lower values widely observed in freshwater largely overlooked in aquatic environments and provide a unified systems. This invariant coupling of N and O isotopic enrichments framework for future investigations of N isotope biogeochemistry. has been shown to originate from fractionation during enzymatic − bond-breakage (8, 21, 22), confirmed directly from in vitro enzyme 2. A Multiprocess Model of NO3 Dual Isotopes: Rationale studies of eukaryotic assimilatory and prokaryotic dissimilatory and Assumptions − − NO3 reductases (11, 23). To evaluate the impact of nitrite reoxidation on coupled NO3 N Interpretation schemes conventionally ascribed to nitrification and O isotope trajectories, termed Δδ18O:Δδ15Nhenceforth,we are also at odds with isotope systematics uncovered in culture devised a time-dependent one-box model simulating the evolution − − observations. Nitrification refers to the sequential oxidation of of N and O isotopologue pools of NO3 and NO2 in a closed − − ammonia (NH3)toNO2 then NO3 by chemoautotrophic bac- system during denitrification coincident with aerobic and anaerobic − teria and archaea, coupled to aerobic respiration. Oxidation of NO2 oxidation by nitrifiers or by anammox bacteria, respectively − − NO2 to NO3 also occurs during the anaerobic oxidation of NH3 (Fig. 1; terms defined in Table 1). The isotopologue-specific for- to N2 by anammox bacteria (24). Nitrification constitutes the sole mulation of the model is described in S2. Equations Used in Time- − − 18 − biological production term for NO3 .TheNO3 δ Ovaluesre- Dependent NO3 Isotope 1 Box Model. − portedly produced by nitrification in freshwater lakes and aquifers In brief, the simulated NO3 pool is influenced by the dissimilative − − span a contended range of 30‰,from−15‰ to 15‰ (2, 25), reduction of NO3 to NO2 (NAR) and the concurrent production − − − attributed to the origin of the oxygen atoms appended to NH3 of NO3 by NO2 oxidation (NXR; Fig. 1). The NO2 pool, in turn, + during the two-step process of nitrification: the biological oxida- reflects the balance of production by NAR and NH4 oxidation − tion of NH3 to NO2 incorporates one oxygen atom from mo- (AMO), as well as oxidation by NXR and reduction to nitric oxide lecular O2 and one from water (7, 26, 27); the subsequent (NIR). Both reductive processes, NAR and NIR, impart normal N − − 15 15 oxidation of NO2 to NO3 incorporates an O atom derived from isotope effects, enar and enir, which are equivalent to their cor- 18 18 water (28). Recent work, however, has revealed kinetic isotope responding O isotope effects, enar and enir (9, 34). During NAR, a 18 effects associated with enzymatic incorporation of each of the branching isotope effect from O atom abstraction, enarBR (19, 29), − 18 − three O atoms into the product NO3 (5, 6), as well an inverse also contributes to the δ OoftheNO2 product. In turn, NXR is − − kinetic isotope effect on the reactant NO2 during oxidation characterized by inverse kinetic isotope effects for NO2 consump- − 15 18 to NO3 (4) and the isotopic equilibration of O atoms between tion, enxr and enxr (4, 5), and a normal isotope effect for O atom − 18 NO2 and water (29, 30). These isotope effects have traditionally incorporation from water, enxr,H2O (5). AMO also has normal − 15 not been considered in interpreting NO3 isotope distributions, isotope effects for N, eamo (3, 41), and for O atom incorporation 18 18 yet play a fundamental role in defining the isotopic composition from H2OandmolecularO2, eamo,H2O and eamo,O2, respectively − + of nitrified NO3 in tandem with compositional differences of the (6). However, we assume no accumulation of ammonium (NH4 ) O atom sources (7, 13, 31). Thus, moving beyond early studies in from the remineralization of organic material, such that there is 18 which the δ ONO3 from nitrification was interpreted as a three- no expression of the N isotope fractionation associated with AMO 15 − 15 part mixture of O atom sources, it is clear that consideration of ( eamo), and the NO2 produced by AMO adopts the δ Nof 15 − several isotope fractionation processes is required for accurate ammonium (δ NNH4). Finally, NO2 is subject to O isotope equil- 18 interpretation of sources and cycling mechanisms. ibration with water having an associated isotope effect, eeq (29, 30). To date, observations in marine systems have revealed linear We further consider the potential role of anaerobic NH3 ox- trajectories of ∼1 and trajectories distinctly above the nominal idation (anammox), which stoichiometrically produces ∼0.26 − − value of 1 associated with water–column denitrification (32–37). moles of NO3 per mole of NH3 oxidized (or per 1.3 moles of NO2 − Positive deviations from 1 are generally interpreted as reflecting reduced) as a metabolic product from the NO2 pool (24, 42). The − the isotopic imprints of biological NO2 reoxidation superimposed model functions as described, except that anammox operates (Fig. 1, − on the trajectory of biological NO3 consumption (34, 38). In blue line) in lieu of aerobic nitrification (Fig. 1, red lines). Notably, 15 + freshwater studies, however, this discrepancy between cultures and unlike in AMO, the δ NoftheNH4 pool does not enter into the − 18 − environment has been scantly acknowledged. Notably, some NO3 mass balance during anammox. Similarly, the δ OofNO3 is − workers have put forth a number of postulates to explain this not indirectly influenced by NO2 production from NH3, which only − conundrum, detailed in S1. Postulates to Explain Deviations from 1 occurs aerobically. NO2 oxidation by anammox is prescribed an 18 15 15 in Δδ O:Δδ N Trajectories in Freshwater. Among these, the bi- inverse N isotope effect, enxrAMX (Table 1) (43). Given no con- − ological production of NO3 by nitrifiers occurring in tandem with straint from culture work the corresponding O isotope effect, 18 denitrification has been proposed to account for the apparent enxrNO2AMX, it is assumed to be of similar magnitude to that for 18 15 − differences in δ Ovs.δ N trajectories between cultures and NO2 oxidizers (Table 1). Model parameter ranges are listed in

E6392 | www.pnas.org/cgi/doi/10.1073/pnas.1601383113 Granger and Wankel Downloaded by guest on September 26, 2021 A equilibration is probably rapid relative to biological cycling given PNAS PLUS the generally low pH (13, 14, 29, 30, 44). In marine denitrifying 18 regions, δ ONO2 measurements also suggest full isotopic equilibration with water (34, 35), despite the higher pH of seawater − (≤8.2). Results for simulations without NO2 isotope equilibration are otherwise presented in the supplements (S3. Simulation − Without Isotope Equilibration of NO2 with Water and Fig. S1). Foremost, results reveal that the evolution of the Δδ18O:Δδ15N trajectory during net denitrification can be substantially deflected from a value of 1 by a co-occurring contribution of newly nitrified − 18 15 NO3 (Fig. 2). Resulting Δδ O:Δδ N trajectories correspond to a broad range of solutions, from 0.1 to 3.2 (in terms of linear fits − relating the dual isotopic composition of initial NO3 to that of any evolved modeled composition), of which a larger share yields slopes below the canonical value of 1. In all cases, the water δ18O emerges as a strong determinant of the Δδ18O:Δδ15N trajectory: the lowest trajectories are associated 18 with the lowest δ OH2O and the highest trajectories with the 18 18 highest δ OH2O (Fig. 2). Mechanistically, the δ Oofnitrified B − 18 NO3 is directly related to the water δ OH2O through both O − isotope equilibration of intermediate NO2 and the incorporation − of an O atom from water during NO2 oxidation. Under condi- − 18 15 tions of full isotopic equilibration of NO2 , Δδ O:Δδ Ntrajec- tories are thereby insensitive to incorporation of O atoms − from water or O2 during any oxidation of NH3 to NO2 .The Δδ18O:Δδ15N trajectories below 1 are associated with relatively δ18 δ18 lower OH2O values that contribute to lower O values for nitrified SCIENCES −

NO3 (Fig. S2). By comparison, the few model solutions in which ENVIRONMENTAL Δδ18O:Δδ15N trajectories are above 1 at standard conditions are as- 18 sociated with higher δ OH2O values, characteristic of marine systems. 18 Conversely, lower δ OH2O values, characteristic of freshwater systems, are largely associated with Δδ18O:Δδ15N trajectories below 1.

− 3.2. Importance of the NO2 Oxidation Flux. Given its importance in − linking the composition of water with that of the evolving NO3 − pool, the extent to which NO3 is produced relative to that reduced (expressed as NXR/NAR) is also clearly an influential Fig. 1. Box model architecture showing N transformations and associated Δδ18 Δδ15 15 18 driver of O: N trajectory. However, this flux is central isotope effects influencing (A) δ NNO3 and (B) δ ONO3. See text and Table 1 − not only because it governs the extent to which nitrified NO3 is for details and references. − added back to the NO3 pool, but also because it modulates the 15 − δ NNO3 returned to the NO3 pool. The influence of NXR/ Section 6 NAR on the Δδ18O:Δδ15N trajectory can be summarized as Table 1 and justified in . For brevity, we also established − − “standard” model conditions having generally midrange values follows (Fig. 2): when NO2 oxidation is nearly equal to NO3 ≥ δ15 for isotope effects (Table 1). reduction (NXR/NAR 0.8), the NNO3 produced by nitrification converges on that removed by denitrification (notwith- − 3. Model Results standing the contribution to the NO2 pool from AMO; S4. − Impact of NO3 Production by AMO), thus nearly “restoring” the We discuss the contribution of distinct aspects of the reaction 15 − 18 δ N of the NO3 pool to its original value. In turn, the δ Oof network, highlighting those features exerting the most influence − − nitrified NO3 is insensitive to NXR/NAR, deriving primarily on the isotope composition of the evolving NO3 pool, namely, 18 − − from the δ OH2O when NO2 is fully equilibrated, such that it the isotope composition of ambient water and the NO2 oxida- 18 can be either higher or lower than the δ ONO3 removed by tion flux. We further explore features that exert moderate in- 18 − denitrification depending on the δ O . Thus, the largest fluence on coupled NO N and O isotope trajectories, including H2O 3 excursions in Δδ18O:Δδ15N trajectories, both above and below 1, isotope effect values, the isotope composition of initial pools, − occur at the higher values of NXR/NAR. In this respect, trajec- and NO2 production via biological ammonia oxidation. We − tories above 1, characteristic of marine denitrifying systems, con- then consider scenarios specific to NO3 production by anam- 18 − gruently arise at elevated δ OH2O. − mox and finally examine the isotope composition of NO2 that When NXR/NAR ratios are low (Fig. 2A), the transient NO2 emerges among model scenarios. Model results are compared − pool is mostly reduced to NO rather than oxidized to NO3 ,ren- Section 4 15 − 15 with environmental observations in . dering the δ NoftheNO2 pool relatively more N-enriched due 15e − 18 to isotopic discrimination by nir.TheNO3 produced from the 3.1. Influence of δ OH2O. Initial tests were parameterized using − 15 oxidation of NO2 is further N-enriched by the inverse effects of standard conditions (Table 1) to explore the sensitivity of the 15 15 − enxr.Therefore,theδ NofnitrifiedNO3 will be greater than Δδ18O:Δδ15N trajectory to the δ18O of ambient water (from Δδ18 Δδ15 − that removed by denitrification, driving the O: Ntrajec- 18 − −20‰ to 0‰), for a range of relative fluxes of NO2 oxidation δ − tories to values below 1. The corresponding OofnitrifiedNO3 (NXR/NAR from 0 to 0.9) and contrasting scenarios of NO2 (again insensitive to the prescribed NXR/NAR ratio due to full − equilibration with water. For simplicity, we predominantly con- isotopic equilibration of NO2 ; Fig. S2) either counters or − 18 15 sider the case of full equilibration of NO2 oxygen isotopes with otherwise augments the negative offset in Δδ O:Δδ Nfrom1 15 − 18 water, which likely characterizes most freshwater systems, where imposed by the δ NNO3 of nitrified NO3 depending on the δ O

Granger and Wankel PNAS | Published online October 4, 2016 | E6393 Downloaded by guest on September 26, 2021 Table 1. Ranges of parameter values explored in the finite-differencing model exercise, including values prescribed in “standard” model parameterizations Initial conditions Description Range Standard Comment Reference

− − [NO3 ]initial Initial [NO3 ]100μmol/L 100 μmol/L − − [NO3 ]initial Initial [NO2 ]0μmol/L 0 μmol/L 15 15 δ NNO3,initial Initial δ NNO3 −10‰ to +15‰ +5‰ 18 18 δ ONO3,initial Initial δ ONO3 −10‰ to +15‰ +5‰ 15 15 δ NNH4,initial Initial δ NNH4 −20‰ to +15‰ +5‰ 18 18 δ OH2O δ O of ambient H2O −20‰ to 0‰ 15 − 18 enar N isotope effect for NO3 reduction +5‰ to +25‰ +15‰ Coupled to enar (9) 18 − 15 enar O isotope effect for NO3 reduction +5‰ to +25‰ +15‰ Coupled to enar (9) 18 − enarBR Branching O isotope effect for NO3 reduction 25‰ 25‰ (19) 15 − 18 enir N isotope effect for NO2 reduction 0‰ to +20‰ +5‰ Coupled to enir (19, 64) 18 − 15 enir O isotope effect for NO2 reduction 0‰ to +20‰ 5‰ Coupled to enir 15 − enxr N isotope effect for NO2 oxidation by nitrifiers −35 to 0‰ −15‰ (4) 18 − enxrNO2 O isotope effect for NO2 oxidation by nitrifiers −7‰ to −3‰ −4‰ (5) or anammox 15 − enxrAMX N isotope effect for NO2 oxidation by anammox −35‰ (43) 15 enxrAMX’ NXR-enabled equilibrium isotope effect between −61‰ (43) − − NO2 and NO3 18 enxrH2O O isotope effect to H2O incorporation by nitrifiers +12‰ to +18‰ +14‰ (5) and anammox 18 − eeq Equilibrium isotope effect between NO2 and H2O 13.5‰ 13.5‰ (30) 15 eamo N isotope effect for NH3 oxidation by aerobic +26‰*26‰*(41) ammonia oxidation 18 eamoH2O O isotope effect for H2O incorporation by aerobic +18‰ to +38‰† 14‰ (6) ammonia oxidation 18 † eamoO2 O isotope effect for O2 incorporation by aerobic +18‰ to +38‰ 14‰ (6) ammonia oxidation − − NXR/NAR Ratio of NO2 oxidation by nitrifiers to NO3 reduction 0–0.9 0.5 by denitrifiers − AMO/NAR Ratio of NO2 production by aerobic ammonia 0–0.9 0 − oxidation vs. by NO3 reduction

*Not expressed given prescription of complete consumption of the ammonium pool. † − The isotope effects of O atom incorporation from O2 and H2O during NH3 oxidation to NO2 have only been determined as a ‘combined’ isotope effect ranging between +18‰ and +38‰ at this time (7). A value of 28‰ was chosen and equally partitioned (+14‰) between the H2O and O2 pools.

− 18 18 15 of this nitrified NO3 (and thus, the δ O of water). Regardless, all Results indicate that the Δδ O:Δδ Ntrajectoryincreases 18 15 15 Δδ O:Δδ N trajectories remain below 1 at low NXR/NAR ratios progressively with increasing enar, the N isotope effect of nitrate 18 A (Fig. S3), even for δ OH2O values characteristic of seawater. reduction (Fig. 3 ). This dynamic is best understood by consid- 15 ering the influence of enar on the composition of the transient 3.3. Influence of N and O Isotope Effects. − − 15 To investigate the leverage NO2 pool subject to reoxidation to NO3 .Theδ NNO2 returned Δδ18 Δδ15 − 15 of specific isotope effects on the evolving O: N trajectories, to the NO3 pool by nitrification is more N-depleted at higher − 15 18 15 we examine the response of NO3 trajectories while holding all enar values, tending to more elevated Δδ O:Δδ Ntrajectories. other parameters constant (standard conditions; Table 1), plotting However, although trajectories above 1 emerge more readily at δ18 15 them against OH2O as a master variable (Fig. 3), given its central higher prescribed enar, the majority of these solutions are not − role in the composition of nitrified NO3 as described above. viable, resulting in observed (or apparent) isotope effects for

NXR/NAR = 0.10 NXR/NAR = 0.50 NXR/NAR = 0.80 A 30 B30 C30 18O 18O 18O 25 water 25 water 25 water 0 ‰ 0 ‰ 0 ‰ 20 -5 ‰ 20 -5 ‰ 20 -5 ‰ -10 ‰ -10 ‰ -10 ‰ 15 -15 ‰ 15 -15 ‰ 15 -15 ‰ -20 ‰ -20 ‰ -20 ‰ 10 10 10 (‰ vs VSMOW) (‰ vs - 3 5 5 5 O NO

18 0 0 0

-5 -5 -5 -5 0 5 10 15 20 25 30 -5 0 5 10 15 20 25 30 -5 0 5 10 15 20 25 30 15N NO - (‰ vs Air) 15N NO - (‰ vs Air) 15N NO - (‰ vs Air) 3 3 3

− 18 18 18 15 15 15 Fig. 2. Predicted evolution of NO3 Δδ ONO3 (δ ONO3 – δ ONO3,initial) plotted on the corresponding Δδ NNO3 (δ NNO3 – δ NNO3,initial) associated with net − denitrification coincident with NO3 production by nitrification. Simulations derive from standard model conditions (Table 1) for incremental prescriptions of 18 − NXR/NAR (A–C) and of δ OH2O (colors), for full oxygen isotopic equilibration of NO2 with water.

E6394 | www.pnas.org/cgi/doi/10.1073/pnas.1601383113 Granger and Wankel Downloaded by guest on September 26, 2021 PNAS PLUS A BC

DEF

− 18 15 − Fig. 3. Predicted NO3 Δδ O:Δδ N trajectories (represented as the apparent linear slope calculated with respect to the initial NO3 isotope composition; − 18 − color scale) associated with net denitrification and coincident NO3 production for a range of δ OH2O given full oxygen isotopic equilibration of NO2 , 15 15 15 15 15 − plotted over (A) enar,(B) enir,(C) enxr, and (D) δ NNO3,initial. Trajectories plotted over enar for NO3 production by anammox only for (E) 30% of total N2 15 15 production by anammox and for (F) 50% N2 production by anammox, with enirAMX = 20‰,and enxrAMX = −35‰. Parameter values are otherwise anchored at standard conditions (Table 1).

15 15 18 − denitrification ( ednf, derived from the change in δ NNO3 relative reduction, enir, also has an apparent influence on the NO3 SCIENCES − 18 15 to the fraction of NO3 removed) substantially higher than the Δδ O:Δδ N trajectory related to its coupling with the corresponding ENVIRONMENTAL ∼ ‰ 15 18 empirical maximum of 30 observed in the environment (Fig. enir (Fig. S4C). Here again, however, any increase in the δ ONO2 − S3). This dynamic arises because the N isotope effects for NO2 imposed by its reduction to NO is ultimately erased due to isotopic 15e − 15e − 18 reduction ( nir), NO2 oxidation ( nxr), and the NXR/NAR equilibration of the NO2 with water, such that the value of enir 15 − 18 15 flux ratio amplify the N-enrichment of the NO3 pool relative is inconsequential to the Δδ O:Δδ N trajectory. 15 18 18 15 to its consumption beyond that imparted specifically by enar, thus In contrast, the amplitude of enxrNO2 does affect the Δδ O:Δδ N 15 18 increasing apparent values of ednf. Thus, viable solutions (i.e., trajectory because the influence of enxrNO2 occurs downstream 15 18 15 − ednf ≤ 30‰)forΔδ O:Δδ N trajectories above 1 emerge of NO2 equilibration. Given this inverse isotope effect, lower 15e ≤ ‰ 18 predominantly at nar amplitudes of 15 . Conversely, viable (more negative) values of enxrNO2 are associated with an in- 15e 18 − solutions stemming from more elevated nar amplitudes are only crease in the δ O of newly nitrified NO3 , and consequently, a − 18 15 produced at relatively low NO3 production (NXR/NAR), cor- relative increase in corresponding Δδ O:Δδ N trajectories Δδ18 Δδ15 18 18 15 responding to O: N trajectories close to 1. Therefore, (Fig. S4E). However, the influence of enxrNO2 on Δδ O:Δδ N Δδ18 Δδ15 18 O: N trajectories above 1, characteristic of marine deni- trajectories is generally muted because enxrNO2 only affects two- ∼ − trifying systems, and of 0.6 for freshwater systems, do not result thirds of the O atoms in the newly nitrified NO3 and because 15e 15 18 at high nar values, suggesting that the organism-level isotope the corresponding enxr has a greater amplitude than enxrNO2 15 18 15 effect for denitrification ( enar) in environmental settings is gen- (Table 1) but opposing influence on Δδ O:Δδ N trajectories. 15 erallynotashighasthe enar of ∼25‰ often observed in culture Finally, the isotope effect associated with the incorporation of Section 4 − 18 conditions (10) ( ). an O atom from water during NO2 oxidation, enxrH2O, also 15e − 18 − The amplitude of nir for the reduction of NO2 also modu- plays a role in determining the δ O of nitrified NO3 . As pre- − 15 18 lates the isotopic evolution of the NO3 , with higher enir (coupled scribed, the δ O of the O atom incorporated from water during 18e Δδ18 Δδ15 − here to nir) corresponding to lower O: Ntrajectories NO2 oxidation is ∼14‰ lower than that of ambient water, thus 15 − 18 (Fig. 3B). Stronger N discrimination during NO2 reduction to contributing a lower δ O than that from the two oxygen atoms δ15 − − 18 NO acts to increase the NoftheNO2 pool, thereby increasing in NO2 . Because the amplitude of enxrH2O has so far proven 15 − − the δ NofNO3 produced from any oxidation of NO2 and to be relatively invariant among cultures and experimental lowering the Δδ18O:Δδ15N trajectory (Fig. S3). strains (5, 7), we do not consider variations of this value on 15e 18 15 The amplitude of the inverse isotope effect prescribed to nxr Δδ O:Δδ N trajectories. 15 also has an important influence on the δ NNO3 of newly nitrified − 15 3.4. Initial Isotope Composition of the NO − Pool. NO3 . Lower enxr values (i.e., more negative) give rise to lower 3 The initial isotopic 15 − δ NNO2, countered by production of correspondingly higher composition of NO3 is also linked to the evolution of its dual iso- 15 18 15 δ NNO3, thereby acting to lower the Δδ O:Δδ N trajectory (Fig. topic trajectory during net consumption coupled with contempora- 15 − − 3C). Conversely, less negative enxr amplitudes produce NO3 neous production. Under conditions of a fully equilibrated NO2 having a relatively lower δ15N, leading to higher Δδ18O:Δδ15N pool, this dependence stems chiefly from the difference between the 18 18 18 trajectories. δ ONO3 and the δ OH2O. Mechanistically, the difference in δ O 18 − In the model, enar is, by design, equivalent to the corresponding values between NO3 and water imposes a corresponding difference 15 18 15 18 − 18 enar. As such, its influence on Δδ O:Δδ N trajectories appears between the δ ONO3 removed by NO3 reduction and the δ ONO3 15e A − − identical to nar (Fig. S4 ). This congruence, however, is only returned to the NO3 pool by NO2 oxidation, in relation to the δ18 − 15 incidental because the ONO2 produced from NO3 reduction is δ NNO3 concurrently being removed and returned. − 15 entirely erased given full isotopic equilibration of NO2 with water. For nearly all combinations of model conditions, the δ NNO3 18 18 Therefore, enar actually has no bearing on the δ Oofnewly added by nitrification is greater than that removed concurrently − − nitrified NO3 . Similarly, the oxygen isotope effect for NO2 by denitrification, owing to the compounding influences of iso-

Granger and Wankel PNAS | Published online October 4, 2016 | E6395 Downloaded by guest on September 26, 2021 − 18 15 tope effects associated with NO2 reduction (normal) and oxi- creases to ∼0.1 and Δδ O:Δδ N trajectories are on the order of 15 15 15 15 dation (inverse), enir and enxr. On its own, the higher δ Nof ∼0.6 for a relatively broader ranges of enar amplitudes (Fig. 3F). − 18 15 nitrified NO3 thus tends to lower the Δδ O:Δδ N trajectory Therefore, if anammox rates in freshwater systems are not strictly − from a value of 1 for most model conditions. bounded by ammonification coupled to NO3 respiration—namely, 18 + The δ ONO3 produced by nitrification either counters or if NH4 is also released via de-sorption from clays (47, 48) or from otherwise amplifies the negative deviations from a Δδ18O:Δδ15N organic matter degradation pathways fueled by fermentation or 15 — − trajectory of 1 imposed by the corresponding δ NNO3. When the other oxidants [e.g., Fe(III), Mn(III/IV)] the production of NO3 18 18 18 Δδ18 Δδ15 initial δ ONO3 is nearer to the δ OH2O, the δ ONO3 produced by anammox could easily account for O: N trajectories of during nitrification is higher than that removed by denitrifica- 0.6 commonly observed in freshwater systems. 18 15 15 tion, such that the Δδ O:Δδ N trajectory increases (Fig. 3D). The amplitude of enxrAMX is subject to some uncertainty. A 18 Conversely, when the initial δ ONO3 is elevated relative to the recent study showed that enrichment cultures of anammox pro- 18 18 − δ15 ∼ ‰ δ OH2O, denitrification removes a δ ONO3 that is similar to, or duced NO3 with a NNO3 61 greater than corresponding 15 greater than, that added by nitrification, contributing further to a δ NNO2 following initial resuspension of cells (interpreted as an 18 15 − decrease in Δδ O:Δδ N trajectory from a value of 1 imposed by “enzyme-catalyzed equilibrium isotope effect” between NO3 and 15 − − corresponding δ NNO3 dynamics. NO2 ), after which NO3 production was associated with a 18 − 15e − ‰ − The difference in δ O between water and NO3 thus pro- nxrAMX amplitude of 35 (43). Accordingly, if NO3 provides a first order rule of thumb to explain the magnitude of duced by anammox had a δ15N ∼61‰ greater than that of cor- 18 15 − − the Δδ O:Δδ N trajectory. For instance, in marine systems, the responding NO2 , then an even smaller contribution of NO3 from 18 − Δδ18 Δδ15 difference between the δ O of subsurface NO3 (≥2‰) and sea- anammox would be required to influence O: Ntrajecto- 18 18 water (∼0‰) is small. At any value of enar,theδ Oofnitrified ries (Fig. S7). From any perspective, anammox clearly emerges − 18 NO3 is greater than the δ ONO3 removed by denitrification, thus as a compelling candidate to explain negative deviations from generating Δδ18O:Δδ15N trajectories above 1 for numerous model Δδ18O:Δδ15N trajectories of 1 observed in freshwater systems. conditions (Fig. 3D and Fig. S5). In contrast, in freshwater systems, Notably, Δδ18O:Δδ15N trajectories remain below 1 regardless of 18 − 15e the difference between the δ OofNO3 and water can be large anammox rate or nxrAMX values, owing to the relatively small 18 18 − (e.g., δ ONO3 of +20‰ and δ OH2O of −10‰ would not be NO2 oxidation flux permitted by anammox stoichiometry. By 18 − Δδ18 Δδ15 unusual), such that the δ OofnitrifiedNO3 tends to be similar to, itself, anammox thus fails to reproduce O: N trajectories 18 or can be lower than, the δ ONO3 removed by denitrification, above 1 observed in marine systems. In our model, even if 18 15 promoting Δδ O:Δδ N trajectories below 1 under most conditions anammox accounts for 100% of N2 production, the corresponding (Fig. 3D and Fig. S5). In this light, Δδ18O:Δδ15N trajectories above NXR/NAR ratio amounts to a mere 0.23, a value insufficient to a nominal value of 1 have an intrinsically greater likelihood of generate Δδ18O:Δδ15N trajectories akin to those observed in ox- 15 ygen deficient zones, regardless of prescribed enxrAMX ampli- emerging in marine systems, whereas trajectories below 1 are − tudes (Fig. S8). Thus a significant input of NO3 from aerobic more likely for freshwater systems. − − NO2 oxidation (or NO2 oxidation coupled to other electron − 18 15 3.5. Influence of NH3 Oxidation to NO2 . As parameterized, the pro- acceptors) is required to explain Δδ O:Δδ N trajectories >1in − 15 duction of NO2 by AMO yields invalid ednf values (>30‰)at oceanic oxygen deficient zones (Section 4.1). − elevated NO2 oxidation fluxes (NXR/NAR ratios ≥ 0.5) while 18 15 − exerting little influence on Δδ O:Δδ N trajectories with lower 3.7. NO2 Isotope Dynamics. Recent methodological advances en- − − δ15 δ18 NO2 oxidation (Fig. S6). For these reasons, we relegate detailed able the direct quantitation of NO2 Nand O at environ- − discussion of the impact of AMO to S4. Impact of NO3 mental concentrations (49, 50), providing a complementary tracer Production by AMO. from which to diagnose environmental N cycling. Under condi- 15 − tions examined here, the predicted δ NoftheNO2 pool ranges − − ‰ δ15 Δδ15 = 3.6. NO3 Production by Anammox. To assess whether NO3 pro- from 30 lower than the corresponding NNO3 ( NNO2-NO3 15 15 duction by anammox is potentially sufficient to induce departures δ NNO2 − δ NNO3 = −30‰), to 10‰ greater than the corre- 18 15 15 18 15 of Δδ O:Δδ N trajectories from a value of 1, we consider two sponding δ NNO3 (Fig. 4). In the context of observed Δδ O:Δδ N 15 scenarios, the first in which anammox rates are bounded by the trajectories, δ NNO2 emerges as sensitive to the prescribed organic matter remineralization stoichiometry of denitrification, NXR/NAR flux ratio (Fig. 4 and Fig. S3). Relatively large neg- 15 18 15 and the second in which anammox rates contribute an equal ative Δδ NNO2-NO3 offsets correspond to elevated Δδ O:Δδ N proportion of the N2 flux relative to denitrification (45). In the trajectories and elevated NXR/NAR amplitudes. Conversely, more + 15 first scenario, anammox rates are controlled by NH4 supplied modest Δδ NNO2-NO3 values are more prevalent at lower NXR/NAR only from the decomposition of organic material coupled with amplitudes (≤0.2) and Δδ18O:Δδ15N trajectories below 1. The − + 18 − NO3 respiration, wherein the release of 1 mol of NH4 from δ O , in turn, can provide affirmation that NO is equil- − NO2 2 organic matter requires respiration of 5.9 mol of NO3 to N2 (46). ibrated with water, which we anticipate in freshwater systems, and + − Each mole of NH4 then reacts with 1 mol NO2 (produced which has been observed in most denitrifying marine systems (34, − transiently by denitrification) to form N2 (24, 42). Concurrently, 35). Therefore, if measured concurrently with those of NO3 (and − − 15 18 − an additional 0.3 mol NO2 is returned to the NO3 pool (42). Of ambient water), the δ Nandδ OofNO2 offer added quanti- − − the total N2 produced, 30% originates from anammox and the rest tative constraints on the relative flux of NO returned to NO − 2 3 from canonical denitrification. The oxidation of NO2 by anam- pool under denitrifying conditions (29, 30, 33, 34). mox therein corresponds to a relatively small NXR/NAR flux ratio of 0.05. 4. Links to Environmental Studies The Δδ18O:Δδ15N trajectories generated from these model 4.1. Marine Systems. Discrete lines of evidence suggest that a δ18 E − conditions are consistently below 1, regardless of OH2O (Fig. 3 ). substantial fraction of the NO2 produced by denitrification in − In this respect, trajectories as low as 0.6, characteristic of freshwater low oxygen zones of the ocean is returned to the NO3 pool by 15e − system, are only attained with prescriptions of relatively low nar of aerobic NO2 oxidation (34, 51, 52). Mechanistically, the redox ≤5‰. Thus, within this scenario, anammox can only explain ex- potential of waters bounding oxygen-deficient zones may be cursions in freshwater Δδ18O:Δδ15N trajectories assuming specifi- dynamic, responding to periodic advection of oxygenated waters 15 − cally low values of enar. near redox boundaries (51, 53). Evidence that NO2 reoxidation When the anammox rate is parameterized to account for 50% occurs concurrently with denitrification originates in part from 18 15 of N2 production, the corresponding NXR/NAR flux ratio in- observations of Δδ O:Δδ N trajectories exceeding an expected

E6396 | www.pnas.org/cgi/doi/10.1073/pnas.1601383113 Granger and Wankel Downloaded by guest on September 26, 2021 − PNAS PLUS 0‰), complete oxygen isotope equilibration of NO2 with sea- − water, and no significant input of NO2 from AMO. The interpretations of recent observations in marine denitrifying waters generally corroborate the predictions from the model presented here. For illustration, we adapted the model architecture outlined above to use an inverse approach to numerically − optimize NO3 isotope measurements along a subsurface isopycnal at the Costa Rica Dome (35) (Fig. 5). The measurements − evidenced a progressive NO3 decrease along the isopycnal surface with a concurrent increase in δ15Nandδ18O, corresponding to a Δδ18O:Δδ15N trajectory >1 and an apparent isotope effect, 15 15 15 15 ednf,of28‰.Valuesof enar, enir, enxr,andNXR/NARwere numerically optimized, iteratively finding the least squares fit to − 15 18 measured NO3 concentration, δ Nandδ O, assuming negligi- − − ble NO2 production by AMO and full equilibration of NO2 with water (S5. Model Inversion). Accordingly, the apparent Δδ18O:Δδ15 N trajectory corresponds to an elevated NXR/NAR flux ratio of 15 15 0.64 ± 0.07, and diagnoses of moderate values for enar, enxr, 15 and enir of 14.1 ± 2.3‰, 10.9 ± 4.4‰, and −16.0 ± 4.5‰, respectively (Fig. 5). Interestingly, the curvature in Δδ18O:Δδ15N trajectories apparent in our model results (Fig. 2) is substantiated by analogous − curvature in dual NO3 isotope trajectories documented in δ15 δ15 Δδ15 − Fig. 4. Predicted difference between NNO2 and NNO3 ( NNO2-NO3)as- – − denitrifying ocean waters (34 37) (Fig. 5). The increase of NO3 sociated with net denitrification and coincident NO3 production plotted against 15e N and O isotope ratios associated with net denitrification along the NXR/NAR ratio, in relation to dnf (color scale). Discrete simulations derive isopycnals in the subsurface follows apparent Δδ18O:Δδ15N tra- from randomized parameter ranges (Table 1), with full oxygen isotopic exchange − = δ18 = ‰ δ15 = ‰ δ18 = jectories distinctly above a nominal slope of 1 that show clear SCIENCES of NO2 , AMO/NAR 0, OH2O -10 , NNO3,initial 5 ,and ONO3,intial −

15 ENVIRONMENTAL 0‰. Open symbols correspond to model solutions where ednf > 30‰. curvature at higher extents of net NO3 consumption. In our model, the degree of curvature increases in proportion to the − 15 − NO2 oxidation flux (Fig. 2), because the δ N of nitrified NO3 δ15 δ18 − − value of 1. Measurements of the N and OofNO3 and increases progressively as a function of net NO3 consumption, − δ18 − NO2 from the Eastern Tropical Pacific and the Peru Upwelling whereas the corresponding O of nitrified NO3 remains di- − δ18 region suggest that NO2 reoxidation is the more likely cause of rectly dependent on the O of water (Fig. S2). The curvature in – δ15 − Δδ18O:Δδ15N trajectory in our model simulations thus appear such patterns (32 34). Specifically, the NofNO2 at the − − 15 validated by analogous patterns in NO3 isotope distributions in subsurface is 30–40‰ lower than the coincident NO3 δ N, reaching δ15N values as much as 60‰ lower than the coexisting marine denitrifying environments. − 15 Anammox has been detected in oxygen minimum zones of the NO3 (Δδ NNO2-NO3 = −30 to −60‰). Best fit solutions to an − ocean, where its potential contribution to the total N2 flux has inverse finite-difference model tracking the evolution of NO3 − − generated considerable debate (54–57), with estimates ranging and NO2 isotopes along isopycnals indicated substantial NO2 from 30% to 100% of the total N2 flux. The model exercise here reoxidation flux, with NXR/NAR ≥ 0.50 (34). Model fits further − suggests that NO3 production by anammox cannot, by itself, 15e ∼ ‰ 15e 18 15 diagnosed moderate nar values of 13 , nxr values on the explain Δδ O:Δδ N trajectories exceeding 1, given diagnoses of − ‰ − 15e − − order of 30 , no fractionation during NO2 reduction ( nir of NO2 reoxidation to the NO3 pool that far exceed stoichiometric

18 15 − 15 18 Fig. 5. (A) Best inverse fit Δδ O:Δδ N trajectory describing NO3 δ N and δ O in an oceanic oxygen deficient zone in the Eastern Tropical North Pacific − 18 (35). Initial NO3 concentrations were interpolated based on the “initial” profiles outside the oxygen deficient zone (35), δ OH2O = 0‰, and the AMO flux is 15 15 15 assumed to be negligible. Best-fit estimates of enar, enir, and enxr are +14.1 ± 2.3‰, +10.9 ± 4.4‰, and −16.0 ± 4.5‰, respectively, with a diagnosed NXR/ 15 15 NAR of 0.64 ± 0.07. Model-predicted Δδ NNO2-NO3 values are between −20‰ and −22‰ and the apparent isotope effect, ednf,is28‰.(B) Best inverse fit 18 15 − 15 18 15 15 15 Δδ O:Δδ N trajectory for NO3 δ N and δ O in a contaminated aquifer in Cape Cod, MA (54). Best-fit estimates of enar, enir, and enxr for the ETNP 15 are +10.2 ± 1.3‰, +9.6 ± 4.2‰, and −24.0 ± 7.7‰, respectively, with a diagnosed NXR/NAR of 0.31 ± 0.08, model-predicted Δδ NNO2-NO3 values between −6‰ 15 and −7‰, and an apparent isotope effect, ednf,of16‰.

Granger and Wankel PNAS | Published online October 4, 2016 | E6397 Downloaded by guest on September 26, 2021 constraints of anammox. Even if anammox is assumed to account served in freshwater systems can be explained by superimposing − for 100% of N2 production, the corresponding NO2 flux remains the isotopic systematics of denitrification with those of oxidative insufficient to generate Δδ18O:Δδ15N trajectories above 1. − − − NO3 production. This dynamic could arise on the premise that Accordingly, analyses of NO3 and NO2 isotope ratios in other redox conditions in aquifers may be dynamic, owing to in- denitrifying regions of the Pacific and Indian Oceans have tercalated microzonation within sediments (60) and/or peri- converged on similar interpretations, ultimately diagnosing a − − odic downward percolation of oxygenated waters. However, substantial return flux of NO2 to the NO3 pool catalyzed nitrifying organisms and their biogeochemical activity are rarely – largely by aerobic nitrification (34 37). detected at the heart of denitrifying aquifers (61, 62). Thus, in the Despite general agreements between model and observations, absence of any O2-requiring transformations, it is likely that any δ15 − − some aspects remain puzzling. For one, the NNO2 in deni- return of NO to the NO pool in anaerobic aquifers is asso- trifying marine waters can be substantially lower than predicted 2 3 15 ciated with anammox, the anaerobic oxidation of NH3 coupled to by the current exercise, with Δδ N values ranging be- − − NO2-NO3 reduction of NO , which yields NO (as well as N ) as a met- tween −30‰ and −60‰ in situ (34, 35), compared with −20% 2 3 2 and −30‰ in our model (Fig. 4). Indeed, the best-fit parameters abolic product (24, 42). Indeed, recent studies suggest that a to the Costa Rica Dome data predict corresponding Δδ15N substantial fraction of N2 production in aquifers originates from NO3-NO2 – values on the order of −20% to −22‰ (Fig. 5) compared with anammox (63 65), rivaling N2 production by canonical de- 15 nitrification in some instances (45). measured Δδ NNO3-NO2 values between −20‰ and −50‰ 15 We turn to a well-studied site of groundwater contamination (35). Lower Δδ NNO3-NO2 values can only be generated in the on Cape Cod, MA (47), to estimate isotope effects and relative current model framework by assuming an elevated reoxidation to − − reduction flux (NXR/NAR), as well as large amplitude isotope NO2 oxidation rates that could explain NO3 isotope distribu- 15 15 ∼ μ − effects for enar and enxr on the order of 30‰ and −30‰, tions therein. Within the contaminant plume, 180 MNO3 de- 15 ∼ μ respectively. However, while realizing Δδ NNO2-NO3 values as low creased to 35 M along an anoxic groundwater flow path. 18 15 − as −60‰ and Δδ O:Δδ N trajectories above 1, such simulations Concomitant with this decrease in NO3 concentration, pro- 15 15 18 result in apparent values of ednf that far exceed the empirical nounced changes in δ NNO3 and δ ONO3 were also observed, 15 limit of 30‰ (Fig. 4). Otherwise, enxr may be effectively sub- increasing from ∼+15‰ to +45‰ and from −1‰ to +18‰, sumed into an equilibrium isotope effect of −61‰, the purported respectively, yielding a corresponding Δδ18O:Δδ15N trajectory of − enzyme-catalyzed equilibrium isotope effect between NO2 and ∼ − − − 0.73 and an apparent N isotope effect for the observed de- 15 NO3 (43), such that the net production of NO3 by the NO2 nitrification, ednf,of∼16‰ (47). We use the inverse approach Δ15 15 15 15 oxidoreductase enzyme can generate NNO2-NO3 values of outlined above to numerically optimize values of e , e , e , ∼− ‰ nar nir nxr 60 , albeit at elevated NXR/NAR flux ratios. This scenario, and NXR/NAR, iteratively finding the least squares fit to measured 15e − however, still results in apparent dnf values that far exceed NO concentration, δ15Nandδ18O(S5. Model Inversion). We also 30‰, thus inconsistent with observations. Arguably, isotopically 3 − make a simplifying assumption that AMO is negligible in the anoxic catalyzed enzymatic isotope equilibration by NO2 oxidoreductase portion of the aquifer. Given the pH and δ18O of groundwater at need not be associated with a net oxidative flux. Such a scenario is this site (∼6.5 and −6.5‰, respectively) (47), we also assume that analogous to simulations here where NXR is nearly equal to NAR − − − any NO2 will be fully equilibrated with water. (with NAR representing NO3 reduction by NO2 oxidoreductase − As expected, deviation of the dual isotopic composition from a in lieu of NO reductase). The equilibrium isotope effect 3 Δδ18O:Δδ15N trajectory of 1–0.73 can be accommodated by con- of −61‰ is then implicitly the sum of the oxidative and reductive − − − comitant production of NO3 along this flow path (Fig. S8). Best- isotope effects with respect to NO2 , −31‰ for NO2 oxidation − − 15 15 15 fit solutions yield estimates of enar, enir,and enxr of 10.6 ± vs. −30‰ for NO2 production from NO3 , wherein the resulting 15 − 0.2‰,10.2± 3.9‰,and−31.3 ± 7.7‰, respectively. Consistent Δ NNO2-NO3 of the NO2 pool is ∼−60‰. Again, apparent val- 15 15e ues of e emerging from such simulations are unrealistic (on with expectations, lower actual values of nar are diagnosed (i.e., dnf 15e ∼ ‰ δ15 the order of 60‰). Thus, within the current model framework, below the observed dnf of 16 ), as the increase in NNO3 is − δ15 the putative enzyme-catalyzed equilibrium isotope effect between also influenced by NO3 production with a relatively high Nvia − − − NO2 and NO3 does not appear to readily explain the very low NXR. Under these conditions, the required amount of NO3 15 ± Δδ NNO2-NO3 observed in marine denitrifying systems. production relative to its removal is on the order of 13 0.3% 15 The discrepancy between our modeled Δδ NNO2-NO3 and (i.e., NXR/NAR = 0.13), much lower than that predicted by re- observations may otherwise derive in part from the constant cent dual isotope models of ocean denitrifying zones in which the N transformation rates in our simulations, as more negative diagnosed NXR/NAR can be as high as 0.9 (34). Although not 15 δ NNO2 values could be generated within representative model − measured in the original aquifer study, these best-fit parameters 15 −— conditions given time/space-variable rates of NO2 accumulation also allow us to predict the δ N of coexisting NO2 for which we and depletion and/or isotope effects. Nevertheless, inverse fits 15 arrive at Δδ NNO2-NO3 values between −6.1‰ and −4.4‰.This allowing for time-variable rates of N transformations still im- prediction of a relatively small δ15N difference between coexisting 15e − ‰ − − plicate nxr values on the order of 30 (34, 35), distinctly pools of NO and NO under aquifer conditions provides a – ‰ 3 2 lower than values observed in nitrifier cultures of 15 (5, 7). target for future studies and a means for refining our estimates of Additionally, the inverse-model fits are contingent on the di- 15e N cycling in groundwater and other environments. minished expression of nir (5, 7), a diagnosis that is also ech- 15e ∼− ‰ 15 Notably, the estimated value for nxr of 31 for the oed in our model (Fig. S3). Such low e amplitudes appear − nir aquifer is higher than that observed in cultures of NO oxidizing contradictory to expectations from cultures and enzymatic 2 bacteria (7), yet largely consistent with that for anammox (43), studies (19, 43, 58, 59) and are also inconsistent with some recent 15 potentially implicating anammox as an important N removal field observations (36), where an increase in δ NNO2 along isopycnal surfaces in a denitrifying eddy at the Peru margin (in process in this aquifer. The diagnosis of NXR/NAR ratio of 0.13 − 15 corresponds to ∼65% of total N2 production in the aquifer which no NO3 remained) was associated with an apparent enir 15 fueled by anammox, a value also consistent with recent in- of 12‰. The very low δ NNO2 observed in marine denitrifying dependent estimates made in the same aquifer (45). Indeed, zones thus remains perplexing and merits further inquiry. − given the stoichiometric constraints of NO3 production by 4.2. Freshwater Systems. As demonstrated here, the empirical anammox, we suggest these model results may prove diagnostic Δδ18O:Δδ15N trajectories between 0.5 and 0.8 recurrently ob- of the role of anammox in ground waters globally.

E6398 | www.pnas.org/cgi/doi/10.1073/pnas.1601383113 Granger and Wankel Downloaded by guest on September 26, 2021 5. Summary and Conclusions ronmentalextentofthepurported“enzyme-catalyzed equilibrium PNAS PLUS ” − − Foremost, our results demonstrate that deviations from the ca- Nisotopeeffect between NO3 and NO2 during anammox nonical Δδ18O:Δδ15N trajectory of 1 for denitrification must should also be further explored. Diagnostic estimates of NXR/ − NAR flux ratios are sensitive to the broad potential range of N emerge due to concurrent NO3 production catalyzed by nitri- − − fication and/or anammox, not only in marine systems, but also in isotope effects during NO2 oxidation to NO3 . Third, the ap- δ15 freshwater systems, where this tenet has been given limited con- parent discrepancy between model and measured NNO2 in − marine systems merits examination to assess the involvement of sideration. Assuming that full O isotopic equilibration of NO2 is − NO2 in potential abiotic/inorganic reactions. Finally, the premise pertinent to both freshwater and marine denitrifying systems, our − 18 results emphasize the sensitivity of the δ O of newly produced that NO2 oxygen isotopes in marine and freshwater denitrifying − 18 NO3 to the δ O of ambient water and the isotope effects for systems are fully equilibrated with water should be further in- O atom equilibration and incorporation. Trajectories above 1 terrogated, as even partial O isotope disequilibrium could influ- ≥ ence diagnostics of N fluxes and associated isotope effects (S3. emerge specifically where the NXR/NAR flux ratio is high ( 0.5) − and predominantly where the difference between the δ18Oof Simulation Without Isotope Equilibration of NO2 with Water). − − NO3 and that of water is small. The diagnosis of high NO3 Continued inquiry into isotope systematics of N cycling will lead to production in the generation of Δδ18O:Δδ15N trajectories above 1 an increasingly robust framework from which to examine N cycle − disputes the notion that anammox could be the sole NO3 -pro- dynamics across ecosystems. ducing reaction in marine denitrifying systems, given stoichio- − 6. Materials and Methods metric and biochemical limitations on the amount of NO2 that − anammox can return to the NO3 pool, and thus implicates a Given the number of variables involved in calculating the model solutions, we − made a number of decisions for model parameterizations to maintain clarity. significant contribution by aerobic NO2 oxidation. Importantly, the majority of solutions to randomized combi- First, we chose a range of isotope effects based on published studies using pure nations of model conditions yield Δδ18O:Δδ15N trajectories <1 cultures of representative organisms and/or purified enzymes when possible − (Table 1). As studies of organism-level isotope effects have demonstrated for simulations using a fully equilibrated NO2 pool. In this re- − broad variability, whether by bacterial strain or growth conditions, for brevity spect, our analysis of NO3 isotopes from a representative and simplicity, we also established “standard” model conditions having gen- Δδ18 Δδ15 aquifer (47) suggests that the characteristic O: N tra- erally midrange values for isotope effects (Table 1). Second, a study published jectory therein (∼0.73) arises from relatively low NXR/NAR flux this year demonstrates that the N and O isotope effects for NIR are coupled −

∼ SCIENCES ratios ( 0.13), which is well aligned with the stoichiometric and differently depending on the type NO2 reductase involved (59). Our simula- 15e 18e biochemical constraints of anammox. Thus, in addition to the tions, however, are based on parameterization of nir and nir coupled in a ENVIRONMENTAL 18 ratio of 1:1, assumed before this insight. Regardless, this inaccuracy does not leverage of distinctive δ OH2O between freshwater and marine − ultimately impact the conclusions reached herein (Section 3.3 and S3.2. Influ- systems, characteristic NO3 isotope trends between denitrifying 18 18 ence of «nar and «nir). Third, the N and O isotope effects for NXR in the groundwater and marine systems also appear to reflect funda- − mental differences in key N transformation pathways operative model parameterization remain uncoupled because studies of NO2 -oxidizing organisms have evidenced no connection in their magnitudes (5). Finally, we in these systems. − δ15 − allow the initial N and O isotopic composition of NO3 to vary across a rep- Within this model framework, the NofNO2 provides an resentative range (−10‰ to +15‰ for both δ15Nandδ18O) while using values additional diagnostic to estimate the relative contribution of + ‰ − of 5 for both as standard conditions. nitrification (and/or anammox) to the NO3 pool. In particular, To date, no information exists on O isotope systematics of anammox, such δ15 − − very deplete NNO2 values relative to the NO3 pool may be that we assume O isotope effects similar to that of NO2 oxidizing bacteria, as characteristic of elevated NXR/NAR flux ratios: a dynamic that enzymatic pathways are analogous (66). However, the organism-level in- 15 − appears corroborated by analogously low δ NNO2 values in verse N isotope effect reported for anammox cultures for NO3 production driven by NXR exhibits higher values (−35‰) (43) than those observed in marine denitrifying zones, albeit lower than those predicted by − cultures of bacterial NO2 oxidizing bacteria (−13‰) (4) and may also include our model simulations. − a very large enzymatically driven isotope equilibrium between NO3 and Some uncertainties also remain that require resolution to en- − NO of −61‰ (43). sure accuracy of interpretations. For one, O isotope effects asso- 2 − ciated with NO3 production by anammox remain undocumented; − ACKNOWLEDGMENTS. This manuscript benefited from comments by two whereas dynamics may be similar to those for NO2 oxidizing anonymous reviewers. This work was supported by National Science bacteria, this should be confirmed. Second, the nature and envi- Foundation Grants EAR-1252089 (to J.G.) and EAR-1252161 (to S.D.W.).

1. Gruber N, Galloway JN (2008) An Earth-system perspective of the global nitrogen 11. Karsh KL, Granger J, Kritee K, Sigman DM (2012) Eukaryotic assimilatory nitrate reductase cycle. Nature 451(7176):293–296. fractionates N and O isotopes with a ratio near unity. Environ Sci Technol 46(11):5727–5735. 2. Kendall C, Elliott EM, Wankel SD (2007) Tracing Anthropogenic Inputs of Nitrogen to 12. Wunderlich A, Meckenstock R, Einsiedl F (2012) Effect of different carbon substrates Ecosystems (Wiley-Blackwell, Malden, MA), pp 375–449. on nitrate stable isotope fractionation during microbial denitrification. Environ Sci 3. Mariotti A, et al. (1981) Experimental determination of nitrogen kinetic isotope Technol 46(9):4861–4868. fractionation: Some principles; illustration for the denitrification and nitrification 13. Snider DM, Spoelstra J, Schiff SL, Venkiteswaran JJ (2010) Stable oxygen isotope ratios processes. Plant Soil 62(3):413–430. of nitrate produced from nitrification: (18)O-labeled water incubations of agricultural 4. Casciotti KL (2009) Inverse kinetic isotope fractionation during bacterial nitrite oxi- and temperate forest soils. Environ Sci Technol 44(14):5358–5364. dation. Geochim Cosmochim Acta 73(7):2061–2076. 14. Kool DM, Wrage N, Oenema O, Van Kessel C, Van Groenigen JW (2011) Oxygen ex- 5. Buchwald C, Casciotti KL (2010) Oxygen isotopic fractionation and exchange during change with water alters the oxygen isotopic signature of nitrate in soil ecosystems. bacterial nitrite oxidation. Limnol Oceanogr 55(3):1064–1074. Soil Biol Biochem 43(6):1180–1185. 6. Casciotti KL, McIlvin M, Buchwald C (2010) Oxygen isotopic exchange and fraction- 15. Böttcher J, Strebel O, Voerkelius S, Schmidt HL (1990) Using isotope fractionation of ation during bacterial ammonia oxidation. Limnol Oceanogr 55(2):753–762. nitrate nitrogen and nitrate oxygen for evaluation of microbial denitrification in a 7. Buchwald C, Santoro AE, McIlvin MR, Casciotti KL (2012) Oxygen isotopic composition sandy aquifer. J Hydrol (Amst) 114(3-4):413–424. of nitrate and nitrite produced by nitrifying cocultures and natural marine assem- 16. Aravena R, Robertson WD (1998) Use of multiple isotope tracers to evaluate de- blages. Limnol Oceanogr 57(5):1361–1375. nitrification in ground water: Study of nitrate from a large-flux septic system plume. 8. Granger J, Sigman DM, Needoba JA, Harrison PJ (2004) Coupled nitrogen and oxygen Ground Water 36(6):975–982. isotope fractionation of nitrate during assimilation by cultures of marine phyto- 17. Amberger A, Schmidt H-L (1987) Natürliche isotopengehalte von nitrat als In- plankton. Limnol Oceanogr 49(5):1763–1773. dikatoren für dessen herkunft. Geochim Cosmochim Acta 51(10):2699–2705. 9. Granger J, Sigman DM, Lehmann MF, Tortell PD (2008) Nitrogen and oxygen isotope 18. Sigman DM, et al. (2001) A bacterial method for the nitrogen isotopic analysis of fractionation during dissimilatory nitrate reduction by denitrifying bacteria. Limnol nitrate in seawater and freshwater. Anal Chem 73(17):4145–4153. Oceanogr 53(6):2533–2545. 19. Casciotti KL, Sigman DM, Hastings MG, Böhlke JK, Hilkert A (2002) Measurement of 10. Kritee K, et al. (2012) Reduced isotope fractionation by denitrification under condi- the oxygen isotopic composition of nitrate in seawater and freshwater using the tions relevant to the ocean. Geochim Cosmochim Acta 92:243–259. denitrifier method. Anal Chem 74(19):4905–4912.

Granger and Wankel PNAS | Published online October 4, 2016 | E6399 Downloaded by guest on September 26, 2021 20. Sigman DM, Lehman SJ, Oppo DW (2003) Evaluating mechanisms of nutrient de- 47. Böhlke JK, Smith RL, Miller DN (2006) Ammonium transport and reaction in con- pletion and C-13 enrichment in the intermediate-depth Atlantic during the last ice taminated groundwater: Application of isotope tracers and isotope fractionation age. Paleoceanography 18(3):1072–1089. studies. Water Resour Res 42:W05411. 21. Shearer G, Schneider JD, Kohl DH (1991) Separating the efflux and influx components 48. Ceazan ML, Thurman EM, Smith RL (1989) Retardation of ammonium and potassium of net nitrate uptake by Synechococcus-R2 under steady-state conditions. J Gen transport through a contaminated sand and gravel aquifer: The role of cation ex- Microbiol 137(5):1179–1184. change. Environ Sci Technol 23(11):1402–1408. 22. Needoba J (2004) Nitrogen Isotope Fractionation by Four Groups of Marine Algae 49. McIlvin MR, Altabet MA (2005) Chemical conversion of nitrate and nitrite to nitrous During Growth on Nitrate (Univ of British Columbia, Vancouver). oxide for nitrogen and oxygen isotopic analysis in freshwater and seawater. Anal 23. Treibergs LA, Granger J (2016) Enzyme level N and O isototope effects of assimilatory Chem 77(17):5589–5595. and dissimilatory nitrate reduction. Limnol Oceanogr, in press. 50. Böhlke JK, Smith RL, Hannon JE (2007) Isotopic analysis of N and O in nitrite and 24. Jetten MSM, et al. (2009) Biochemistry and molecular biology of anammox bacteria. nitrate by sequential selective bacterial reduction to N2O. Anal Chem 79(15): Crit Rev Biochem Mol Biol 44(2-3):65–84. 5888–5895. 25. Mayer B, Bollwerk SM, Mansfeldt T, Hütter B, Veizer J (2001) The oxygen isotope 51. Anderson JJ, Okubo A, Robbins AS, Richards FA (1982) A model for nitrate distribu- composition of nitrate generated by nitrification in acid forest floors. Geochim tions in oceanic oxygen minimum zones. Deep-Sea Res A, Oceanogr Res Pap 29(9): Cosmochim Acta 65(16):2743–2756. 1113–1140. 26. DiSpirito AA, Hooper AB (1986) Oxygen exchange between nitrate molecules during 52. Beman JM, Leilei Shih J, Popp BN (2013) Nitrite oxidation in the upper water column nitrite oxidation by Nitrobacter. J Biol Chem 261(23):10534–10537. and oxygen minimum zone of the eastern tropical North Pacific Ocean. ISME J 7(11): – 27. Hollocher TC (1984) Source of the oxygen atoms of nitrate in the oxidation of nitrite 2192 2205. by Nitrobacter agilis and evidence against a P-O-N anhydride mechanism in oxidative 53. Bettencourt JH, et al. (2015) Boundaries of the Peruvian oxygen minimum zone – phosphorylation. Arch Biochem Biophys 233(2):721–727. shaped by coherent mesoscale dynamics. Nat Geosci 8(12):937 940. 28. Kumar S, Nicholas DJD, Williams EH (1983) Definitive 15N NMR evidence that water 54. Babbin AR, Keil RG, Devol AH, Ward BB (2014) Organic matter stoichiometry, flux, – serves as a source of ‘O’ during nitrite oxidation by Nitrobacter agilis. FEBS Lett 152(1): and oxygen control nitrogen loss in the ocean. Science 344(6182):406 408. 71–74. 55. Ward BB, et al. (2009) Denitrification as the dominant nitrogen loss process in the – 29. Casciotti KL, Böhlke JK, McIlvin MR, Mroczkowski SJ, Hannon JE (2007) Oxygen iso- Arabian Sea. Nature 461(7260):78 81. topes in nitrite: Analysis, calibration, and equilibration. Anal Chem 79(6):2427–2436. 56. Dalsgaard T, Thamdrup B, Farias L, Revsbech NP (2012) Anammox and denitrification 30. Buchwald C, Casciotti KL (2013) Isotopic ratios of nitrite as tracers of the sources and in the oxygen minimum zone of the eastern South Pacific. Limnol Oceanogr 57(5): – age of oceanic nitrite. Nat Geosci 6(4):308–313. 1331 1346. 31. Fang Y, et al. (2012) Low δ18O values of nitrate produced from nitrification in tem- 57. Jensen MM, et al. (2011) Intensive nitrogen loss over the Omani Shelf due to anammox coupled with dissimilatory nitrite reduction to ammonium. ISME J 5(10): perate forest soils. Environ Sci Technol 46(16):8723–8730. 1660–1670. 32. Sigman DM, et al. (2005) Coupled nitrogen and oxygen isotope measurements of 58. Bryan BA, Shearer G, Skeeters JL, Kohl DH (1983) Variable expression of the nitrogen nitrate along the eastern North Pacific margin. Global Biogeochem Cycles 19(4): isotope effect associated with denitrification of nitrite. JBiolChem258(14):8613–8617. GB4022. 59. Martin TS, Casciotti KL (2016) Nitrogen and oxygen isotopic fractionation during 33. Casciotti KL, McIlvin MR (2007) Isotopic analyses of nitrate and nitrite from reference microbial nitrite reduction. Limnol Oceanogr 61(3):1134–1143. mixtures and application to Eastern Tropical North Pacific waters. Mar Chem 107(2): 60. Hartog N, Griffioen J, van der Weijden CH (2002) Distribution and reactivity of O2- 184–201. reducing components in sediments from a layered aquifer. Environ Sci Technol 36(11): 34. Casciotti KL, Buchwald C, McIlvin M (2013) Implications of nitrate and nitrite isotopic 2338–2344. measurements for the mechanisms of nitrogen cycling in the Peru oxygen deficient 61. Harris SH, Smith RL (2009) In situ measurements of microbially-catalyzed nitrification zone. Deep Sea Res Part I Oceanogr Res Pap 80:78–93. and nitrate reduction rates in an ephemeral drainage channel receiving water from 35. Buchwald C, Santoro AE, Stanley RHR, Casciotti KL (2015) Nitrogen cycling in the coalbed natural gas discharge, Powder River Basin, Wyoming, USA. Chem Geol secondary nitrite maximum of the eastern tropical North Pacific off Costa Rica. Global 267(1-2):77–84. Biogeochem Cycles 29(12):2061–2081. 62. Smith RL, Baumgartner LK, Miller DN, Repert DA, Böhlke JK (2006) Assessment of 36. Bourbonnais A, et al. (2015) N-loss isotope effects in the Peru oxygen minimum zone nitrification potential in ground water using short term, single-well injection exper- studied using a mesoscale eddy as a natural tracer experiment. Global Biogeochem iments. Microb Ecol 51(1):22–35. Cycles 29(6):793–811. 63. Clark I, et al. (2008) Origin and fate of industrial ammonium in anoxic ground water - 37. Gaye B, Nagel B, Dähnke K, Rixen T, Emeis KC (2013) Evidence of parallel de- 15N evidence for anaerobic oxidation (anammox). Ground Water Monit Remediat nitrification and nitrite oxidation in the ODZ of the Arabian Sea from paired stable 28(3):73–82. – isotopes of nitrate and nitrite. Global Biogeochem Cycles 27(4):1059 1071. 64. Robertson WD, et al. (2012) Natural attenuation of septic system nitrogen by 38. Wankel SD, Kendall C, Pennington JT, Chavez FP, Paytan A (2007) Nitrification in the anammox. Ground Water 50(4):541–553. euphotic zone as evidenced by nitrate dual isotopic composition: Observations from 65. Erler DV, Eyre BD, Davison L (2008) The contribution of anammox and denitrification Monterey Bay, California. Global Biogeochem Cycles 21(2):GB2009. to sediment N2 production in a surface flow constructed wetland. Environ Sci Technol 39. Wenk CB, et al. (2014) Partitioning between benthic and pelagic nitrate reduction in 42(24):9144–9150. – the Lake Lugano south basin. Limnol Oceanogr 59(4):1421 1433. 66. Kartal B, et al. (2011) Molecular mechanism of anaerobic ammonium oxidation. 40. Wunderlich A, Meckenstock RU, Einsiedl F (2013) A mixture of nitrite-oxidizing and Nature 479(7371):127–130. δ denitrifying microorganisms affects the 18O of dissolved nitrate during anaerobic 67. Böhlke JK, Mroczkowski SJ, Coplen TB (2003) Oxygen isotopes in nitrate: New ref- δ microbial denitrification depending on the 18O of ambient water. Geochim erence materials for 18O:17O:16O measurements and observations on nitrate-water Cosmochim Acta 119:31–45. equilibration. Rapid Commun Mass Spectrom 17(16):1835–1846. 41. Casciotti KL, Sigman DM, Ward BB (2003) Linking diversity and stable isotope frac- 68. Révész K, Böhlke JK (2002) Comparison of delta18O measurements in nitrate by tionation in ammonia-oxidizing bacteria. Geomicrobiol J 20(4):335–353. different combustion techniques. Anal Chem 74(20):5410–5413. 42. Strous M, Kuenen JG, Jetten MSM (1999) Key physiology of anaerobic ammonium 69. Houlton BZ, Sigman DM, Hedin LO (2006) Isotopic evidence for large gaseous nitro- oxidation. Appl Environ Microbiol 65(7):3248–3250. gen losses from tropical rainforests. Proc Natl Acad Sci USA 103(23):8745–8750. 43. Brunner B, et al. (2013) Nitrogen isotope effects induced by anammox bacteria. Proc 70. Frey C, Hietanen S, Jürgens K, Labrenz M, Voss M (2014) N and O isotope fractionation Natl Acad Sci USA 110(47):18994–18999. in nitrate during chemolithoautotrophic denitrification by Sulfurimonas gotlandica. 44. Kool DM, Wrage N, Oenema O, Dolfing J, Van Groenigen JW (2007) Oxygen exchange Environ Sci Technol 48(22):13229–13237. between (de)nitrification intermediates and H2O and its implications for source de- 71. Knöller K, Vogt C, Haupt M, Feisthauer S, Richnow HH (2011) Experimental in- termination of NO3- and N2O: A review. Rapid Commun Mass Spectrom 21(22): vestigation of nitrogen and oxygen isotope fractionation in nitrate and nitrite during 3569–3578. denitrification. Biogeochemistry 103(1):371–384. 45. Smith RL, Böhlke JK, Song B, Tobias CR (2015) Role of anaerobic ammonium oxidation 72. Kaneko M, Poulson SR (2013) The rate of oxygen isotope exchange between nitrate (anammox) in nitrogen removal from a freshwater aquifer. Environ Sci Technol and water. Geochim Cosmochim Acta 118:148–156. 49(20):12169–12177. 73. Brunner B, Bernasconi SM, Kleikemper J, Schroth MH (2005) A model for oxygen and 46. Richards FA (1965) Anoxic basins and fjords. Chemical Oceanography, eds Rily J, sulfur isotope fractionation in sulfate during bacterial sulfate reduction processes. Skirrow G (Academic Press, New York), Vol 1, pp 611–645. Geochim Cosmochim Acta 69(20):4773–4785.

E6400 | www.pnas.org/cgi/doi/10.1073/pnas.1601383113 Granger and Wankel Downloaded by guest on September 26, 2021