Nitrous oxide (N

Nitrous oxide (N

Much of the N

Nonetheless, a significant fraction of the N

The stable natural abundance nitrogen and oxygen isotopes of N

Previous studies have used

Experiments were performed at three stations in the eastern tropical North Pacific on the R/V

Ambient [NO

Locations of the three stations sampled for this study. Stations are plotted on top of World Ocean Atlas oxygen saturation ( %) at 250 m depth (World Ocean Atlas, 2013). Schlitzer, Reiner, Ocean Data View,

Incubation depths were chosen to target prominent hydrographic features: the primary NO

After sample collection, a 2 mL He headspace was created in each bottle by displacing the 2 mL sample from the bottle with He. At most (all but two) anoxic depths at stations PS2 and PS3, samples were sparged with He gas for 90 min at a flow rate of at least 100 mL min

Time series were constructed by sacrificing triplicate bottles over a time course, rather than by resampling the incubation bottles over time. A total of 27 incubation samples were thus produced at each experimental depth, comprised of triplicate samples for each of the three time points and three tracers. For each station and depth, nine samples were amended with

Eight 160 mL glass serum bottles were prepared with a chemiluminescent oxygen optode spot (PyroScience) affixed to the inner glass wall with silicone glue. These bottles were incubated alongside experimental bottles to monitor dissolved [O

The optode [O

Optode [O

Two steps were taken to prepare incubation samples for N

Samples were measured for N

In natural abundance samples, pyisotopomer solves the following four equations to obtain

For

The concentration of

After N

Samples incubated with

Finally, samples incubated with

The rates of NH

A time-dependent model was constructed to infer the rates and mechanisms of N

Schematic of the forward-running model used to solve for rates of N

The concentration of each nitrogen species was modeled as

The pattern of N

To represent each N

To relate the

The model was optimized against isotopocule data at each time step in each tracer experiment (Fig. S4). The parameters being optimized (inputs to the cost function) were the second-order rate constants

To ground-truth the model, rates of N

Station PS1, which was at the edge of the ODZ, represented a “background” station with no secondary NO

NO

Rates of NO

NO

NH

At each station, the observed rates of net

Net

Net

For example, in the secondary NO

At many stations and depths, the net production of

Based on model results, the rates of N

N

Hybrid N

N

The percentage of N

N

The percentage of hybrid N

N

The oxygen dependencies of N

N

Hybrid N

N

Hybrid N

The rate of N

In this study, we found that N

Based on our rate data, N

Hybrid N

Although our data do not allow us to comment directly on the enzymatic machinery of hybrid N

These findings of equal

Simulated values of

The unequal production of

The rates of N

When [O

We found three depths near the surface where hybrid production comprised a smaller percentage (0 %–68 %) of total N

N

These results showed that N

While this study and others have found that hybrid N

The maximum N

N

Care was taken to minimize the effects of experimental setup on the microbial communities in each sample. In addition to the steps taken to prevent oxygen contamination (described in Sect. 2: Methods), a relatively short 24 h incubation period was selected to minimize bottle effects and shifts in the microbial community composition over the course of each incubation. Nonetheless, sample collection, preparation, and incubation conditions could have affected the microbial communities in several ways. Firstly, samples were frequently collected from depths where the water temperature was cooler than that of the laboratory, and, while samples were returned to a cool temperature during incubation (12 °C), they were exposed to warmer temperatures (

Other processes may have contributed to N

We applied N

Based on the equal production of

N

Since only 2 mL of sample was available for preparation and analysis of nitrate isotopes using the denitrifier method, it was not possible to always achieve consistent peak areas. Instead of discarding low-peak-area samples, however, we wanted to establish a method to estimate the uncertainties associated with individual samples based on their peak area. This uncertainty arises from a correction scheme for

In brief, the first step of this method is to calculate the peak area and

In practice, we start with a simple mass balance that states that the measured

Equation (A2) can be expressed as a linear equation,

We can obtain the mean blank peak area

Finally, we obtain

We assign the dummy samples a range of theoretical measured peak areas,

As a first example, we assign all of the theoretical samples the same

We correct the range of

Then we calculate the error associated with each dummy sample using

Following this exercise with a range of theoretical peak areas from 0.5 to 10 Vs produces the following curve (Fig. A1). It shows that these theoretical errors increase as peak area decreases, reflecting the basis of the error.

Repeating this exercise with a range of

Finally, we fit a function of the following form through these theoretical data:

This procedure was repeated for each denitrifier run to produce coefficients

The data reported in this study can be found in the Stanford Digital Repository (

The supplement related to this article is available online at:

CLK and KLC conceptualized the study, with input from CF and BBW. CLK and NMT carried out the experiments at sea, with assistance and supervision from CF and BBW. CLK and PAB analyzed the incubation samples in the laboratory. CLK performed the formal analysis of the data, developed the model code, and performed the model optimizations. XS provided N

The contact author has declared that none of the authors has any competing interests.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors.

This article is part of the special issue “Low-oxygen environments and deoxygenation in open and coastal marine waters”. It is not associated with a conference.

We would like to thank Stanford University and the Stanford Research Computing Center for providing computational resources and support that contributed to this research. We would also like to thank Julie Granger, Scott Wankel, and two anonymous reviewers for their constructive feedback during the review process. This research was supported by US NSF grant no. OCE-1657868 to Karen L. Casciotti. Colette L. Kelly was supported by an NSF Graduate Research Fellowship. The authors declare no competing financial interests.

This research has been supported by the Directorate for Geosciences (grant no. OCE-1657868).

This paper was edited by Nicolas Brüggemann and reviewed by Scott Wankel and two anonymous referees.