Analysis of ¹⁵N-NO₃⁻ Via Anoxic Slurries Coupled to MIMS Analysis: An Application to Estimate Nitrification by Burrowing Macrofauna

Abstract

The increasing use of the stable isotope ¹⁵N-NO₃⁻ for the quantification of ecological processes requires analytical approaches able to distinguish between labelled and unlabeled N forms. We present a method coupling anoxic sediment slurries and membrane inlet mass spectrometry to quantify dissolved ¹⁵N-NO₃⁻ and ¹⁴N-NO₃⁻. The approach is based on the microbial reduction of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ mixed pool, the determination of the produced ²⁹N₂ and ³⁰N_2, and the calculation of the original ¹⁵N-NO₃⁻ and ¹⁴N-NO₃⁻ concentrations. The reduction is carried out in 12 mL exetainers containing 2 mL of sediment and 10 mL of water, under anoxia. To validate this approach, we prepared multiple standard solutions containing ¹⁵N-NO₃⁻ alone or in combinations with ¹⁴N-NO₃⁻, with final concentrations varying from 0.5 to 3000 µM. We recovered nearly 90% of the initial ¹⁴N-NO₃⁻ or ¹⁵N-NO₃⁻, over a wide range of concentrations and isotope ratios in the standards. We applied this method to a ¹⁵N-NO₃⁻ dilution experiment targeting the measurement of nitrification in sediments with and without the burrower Sparganophilus tamesis. The oligochaete did not stimulate nitrification, likely due to limited ventilation and unfavorable conditions for nitrifiers growth. The proposed method is reliable, fast, and could be applied to multiple ecological studies.

1. Introduction

Nitrogen (N) pollution and increasing eutrophication of aquatic environments have increased attention concerning the regulation of the N cycle by microbial and primary producer communities [1]. Different methods and techniques have been developed for measuring N transformations, in particular in sediments, and they are more and more based on the use of the N stable isotope ¹⁵N [2,3,4]. Since the introduction of the isotope pairing technique (IPT) [5], denitrification has been a target process for nearly 25 years, and its role in aquatic ecosystems has been often overemphasized as compared to other microbial transformations [6]. The IPT is based on the addition of ¹⁵N-NO₃⁻ to the water phase of micro or mesocosms and the quantification of the produced ²⁹N₂ and ³⁰N₂. The IPT should not be used when anammox contributes significantly to the N₂ production. Thus, a bioassay for anammox determination should be always performed before using this technique. Denitrification bioassays, besides the accurate determination of N₂ production within sediments, may be used to determine ¹⁵N-NO₃⁻ concentrations in the water phase, allowing experimental evaluations of other processes [7]. ¹⁵N-NO₃⁻ determination, in combination with ¹⁵N-NH₄⁺ analysis, may allow the precise quantification of relevant processes of the N cycle as the dissimilative reduction of nitrate to ammonium (DNRA) or the nitrification rate [8,9,10]. The latter are much less studied processes as compared to denitrification [6]. In recent years, the determination of N isotopes in dissolved N₂ has become fast and accurate thanks to the introduction of membrane inlet mass spectrometry (MIMS, [11]). The membrane inlet mass spectrometer does not require creating a headspace in water samples, which are circulated in a gas-permeable capillary where the dissolved gases are extracted by a vacuum and directed to the mass spectrometer [8,9,11].

Previous methodological reports for the determination of N isotopes in pools of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ were based on bioassays with pure bacterial culture of denitrifiers, the quantification of ²⁹N₂ and ³⁰N₂ and of the ¹⁴N to ¹⁵N ratios [7,8]. Alternatively, ¹⁵N-NO₃⁻ can be accurately determined via the use of strong reductants as cadmium and sodium azide, as the chemical reduction to N₂O and the analysis of the ^14/15N pool in the N₂O [12,13,14]. In this work, we propose a relatively simple, inexpensive and quick method to measure a wide range of concentrations of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ combining anoxic sediment slurries and MIMS analysis. Original ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ concentrations are back calculated according to the equations and assumptions of the isotope pairing technique. We present here results from multiple laboratory tests aiming at the precise calibration of the method and of the specific performances of the sediment we employed (e.g., denitrification capacity, co-occurrence of anammox) under different experimental conditions (e.g., slurry concentrations, incubation time, variable isotopic ratios). The accurate determination of labelled and unlabeled dissolved nitrate has multiple implications in ecological studies of N cycling. In particular, it may offer the possibility to accurately assess the rates of nitrification in intact sediment cores, a process that is understudied as compared to other microbial N transformations. We therefore applied the method to a laboratory experiment targeting the rates of nitrification, calculated via the dilution of ¹⁵N-NO₃⁻, in riverine sediments with and without the burrowing oligochaete Sparganophilus tamesis. Macrofauna, in order to promote aerobic conditions in their burrows and get rid of metabolic waste, pump oxic water within otherwise anoxic and ammonium-rich mud. The activity of bioturbating macrofauna can stimulate microbial nitrification and increase the production of ¹⁴N-NO₃⁻ from the pore water ¹⁴N-NH₄⁺ pool. Such production leads to decreasing ¹⁵N-NO₃⁻ to ¹⁴N-NO₃⁻ ratios in the water column, which are proportional to nitrification rates and can be accurately quantified.

2. Materials and Methods

Instead of using pure cultures of denitrifying bacteria or performing chemical reductions, we added eutrophic riverine sediments, which were collected from a shallow-water (50 cm) wetland area of the Mincio River (Northern Italy), to water samples containing labelled and unlabeled nitrate. Previous experiments suggested that these organic sediments display high rates of denitrification, whereas anammox was never detected [15]. Sediments were muddy, with an average porosity of 0.77 and an organic matter content of 9.2% as loss on ignition. They were collected via plexiglass liners; the upper 5 cm layer was extruded and sieved (0.5 mm sieve) to remove macrofauna and large debris and homogenized. Variable volumes of the sediment homogenate (from 0.5 to 6 mL, see the next paragraphs), together with glass beads, were transferred to 12 mL gas-tight vials with screw cap and rubber septum (Labco Exetainers, Lampeter, UK). The vials were then completely filled with water containing different concentrations of ¹⁵N-NO₃⁻ (K¹⁵NO₃ 98% atom, Cambridge Isotope Laboratories, MA, USA), previously flushed with pure N₂ to remove O₂. All treatments were done in triplicates and always combined with blanks (O₂-free deionized water). Once closed without bubbles, the exetainers were put on a rotating shaker in order to maintain sediments in suspension. They were incubated in the dark and at environment temperature (20 °C); the biological activity was stopped by the addition of ZnCl₂ (200 μL, 7 M). Thereafter, they were centrifuged, maintained at 4 °C and analyzed within two weeks for ²⁹N₂ and ³⁰N₂ by MIMS (Bay Instruments, Easton, MD, USA) at the University of Ferrara (Ferrara, Italy).

With the purpose of experimentally defining the optimal slurry concentration and incubation time and to evaluate the efficiency of the method in terms of ¹⁵N recovery along ¹⁵N-NO₃⁻ concentrations series or variable ¹⁵N to ¹⁴N ratios in the dissolved nitrate pool, we performed four sequential tests, described in paragraphs 2.1 to 2.4 and summarized in

Table 1. Summary of the experimental conditions of each test. Every tested condition was run in triplicate.

		[15N-NO3−] µM			[14N-NO3−] µM			Time (h)		[Slurries] mL L−1
Slurry concentration series		50			0			20		42 83 125 167	208 250 292 333	375 417 458 500
Time series		2 20 200 300			0			0.5 1 1.5 2 2.5 3	5 7 10 22 34 48	167
Concentration series		0.5 2.5 5 10 20 30 40 50	75 100 150 200 300 400 500 600	700 800 900 1000 1500 2000 3000	0			20		167
Quantification of labelled and unlabeled N-NO3−	(a)	0.2 0.6 2.5 6.5 16.5	33.0 50.0 67.0 83.5 93.5	97.5 99.4 99.8	99.8 99.4 97.5 93.5 83.5	67.0 50.0 33.0 16.5 6.5	2.5 0.6 0.2	20		167
	(b)	37.5 40 42.5 45 47.5	50 52.5 55 57.5 60	62.5	62.5 60 57.5 55 52.5	50 47.5 45 42.5 40	37.5

.

2.1. Slurry Concentration Series

In order to choose the appropriate sediment to water ratio of the slurry, we tested different slurry concentrations (from 42 to 500 mL L⁻¹, in triplicate), while we maintained a fixed level of ¹⁵N-NO₃⁻ (50 µM). To this purpose, 12 different sediment to water ratios were analyzed (Table 1). All vials were incubated in the dark for 24 h.

2.2. Time Series Experiment

In order to determine the appropriate incubation time to denitrify the dissolved nitrate, we performed a time series experiment. To this purpose, slurries were realized using four different concentrations of ¹⁵N-NO₃⁻ (from 2 to 1000 µM, in triplicate) that were incubated from 0.5 to 48 h (Table 1). The concentration of the slurry was fixed at 167 mL L⁻¹ (2 mL of sediment and 10 mL of water). At different time intervals, for a total of 13 different incubation times, 3 exetainers per level of ¹⁵N-NO₃⁻ were sacrificed, adding ZnCl₂ (200 μL, 7 M) to stop microbial activity.

2.3. Concentration Series Experiment

To quantify the total denitrification capacity of the microbial community in the sediment homogenate, a series of experiments were realized with increasing concentrations of ¹⁵N-NO₃⁻. To this purpose, we tested 24 different levels of nitrate (from 0 to 3000 µM ¹⁵N-NO₃⁻, in triplicate; Table 1). The concentration of the slurry was fixed at 167 mL L⁻¹ and based on previous findings the vials were incubated in the dark for 20 h.

2.4. Quantification of Labelled and Unlabeled N-NO₃⁻

As in most experimental applications, both labelled and unlabeled nitrate are present in solution, 24 standards containing different ratios of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ (from 0.2 to 99.8 atom% of the isotope) were prepared in triplicate (Table 1). Regardless of the ratios of the two isotopes, the final nitrate concentration was always 100 µM. The concentration of the slurry was fixed at 167 mL L⁻¹, and the vials were incubated in the dark for 20 h.

2.5. Anammox Assay

In order to exclude the occurrence of anammox, which would introduce an error in the IPT calculations [16], we checked whether the production of ²⁹N₂ followed the trend expected by the binomial distribution of ²⁸N₂, ²⁹N₂ and ³⁰N₂ due to denitrification alone. To this purpose, we made a series of slurries (in triplicate) with variable ¹⁵N-NO₃⁻ concentrations (from 5 to 200 µM) and a fixed ¹⁴N-NO₃⁻ concentrations (50 µM). The occurrence of anammox would result in additional ²⁹N₂ production from the combination of ¹⁴NH₄⁺ and ¹⁵N-NO₂⁻.

2.6. Calculations

The concentration of the total unlabeled N₂ (TN₂) present in the samples was calculated from the ²⁸N₂ signal of the MIMS. The concentrations of ²⁹N₂ and ³⁰N₂ (µM) were calculated from Equations (1) and (2):

$$\left[{}_{}{}^{29}\mathrm{N}{}_2\right]= \frac{{}_{}{}^{29}\mathrm{N}{}_2}{{}_{}{}^{28}\mathrm{N}{}_2} \times \left[{\mathrm{TN}}_2\right] ,$$

(1)

$$\left[{}_{}{}^{30}\mathrm{N}{}_2\right]= \frac{{}_{}{}^{30}\mathrm{N}{}_2}{{}_{}{}^{28}\mathrm{N}{}_2} \times \left[{\mathrm{TN}}_2\right] .$$

(2)

where ²⁹N₂/²⁸N₂ and ³⁰N₂/²⁸N₂ are the excess of ²⁹N₂ and ³⁰N₂ over the total ²⁸N₂, measured as mass 29 and 30 to 28 ratios. The excess was calculated correcting measured ratios by the natural background values.

The concentration of the ²⁸N₂ originated via denitrification of ¹⁴N-NO₃⁻ was calculated from the Equation (3), assuming a binomial distribution of the produced ²⁸N₂, ²⁹N₂ and ³⁰N₂ [5]:

$$\left[{}_{}{}^{28}\mathrm{N}{}_2\right]= \frac{{\left(\frac{\left[{}_{}{}^{29}\mathrm{N}{}_2\right]}{2}\right)}^2}{\left[{}_{}{}^{30}\mathrm{N}{}_2\right]}$$

(3)

The concentrations of ¹⁵N-NO₃⁻ and ¹⁴N-NO₃⁻ (µM) were then calculated with Equations (4) and (5):

$$\left[{}_{}{}^{14}\mathrm{N}-{\mathrm{NO}}_3^{-}\right]= \left[{}_{}{}^{29}\mathrm{N}{}_2\right] +2 \times \left[{}_{}{}^{28}\mathrm{N}{}_2\right] ,$$

(4)

$$\left[{}_{}{}^{15}\mathrm{N}-{\mathrm{NO}}_3^{-}\right]= \left[{}_{}{}^{29}\mathrm{N}{}_2\right] +2 \times \left[{}_{}{}^{30}\mathrm{N}{}_2\right] .$$

(5)

2.7. Application of the Method to a ¹⁵N-NO₃⁻ Dilution Experiment to Measure Nitrification Rates in Bioturbated Sediments

2.7.1. Microcosm Set Up and Incubation

We applied the proposed method in order to quantify the rates of nitrification in bare and bioturbated sediments (Figure 1 and Figure 2). Nitrification can be assumed as the sum of the nitrate produced within sediments and diffusing to the water column (¹⁴N-NO₃⁻-efflux) and of the nitrate produced within sediments, diffusing to the anoxic layers and denitrified (D_N). Nitrate efflux can be calculated by means of dilution of ¹⁵N-NO₃⁻ in the water phase [17], while coupled nitrification–denitrification can be calculated by applying the IPT [5]. The calculations proposed here would lead to underestimated nitrification activity in the presence of significant NO₃⁻ uptake by primary producers at the sediment–water interface or of high rates of DNRA.

In order to measure nitrification rates in bioturbated sediment, nearly 5 L of sediment was collected from the same area previously described and sieved to remove large debris and macrofauna, then homogenized. The sediment homogenate was transferred into cylindrical plexiglass microcosms (n = 16, inner diameter 8 cm, height 10 cm) as described in [18]. After one day of acclimatization, individuals of the oligochaete S. tamesis were added in 8 microcosms to a final density of 800 ind m⁻² (high density treatment H), reflecting in situ values [19]. All oligochaetes burrowed within a few minutes into the sediments. The remaining 8 microcosms, without macrofauna, were used as controls (C). Microcosms C and H were preincubated for one week in large aquaria containing well mixed and aerated in situ water (NH₄⁺ 1.8 ± 0.2 µM, NO₃⁻ 103.7 ± 3.6 µM; NO₂⁻ 3.1 ± 0.1 µM) maintained at 20 °C. After the preincubation, 8 microcosms (4 with bare sediments and 4 with oligochaetes) were included in liners with a water phase of 16 cm, mixed by Teflon-coated stirring bars rotating at 60 rpm and with the top open, in order to ensure oxic conditions (Figure 1). Each liner was added with 1.8 mL of 15 mM ¹⁵N-NO₃⁻ solution to have a final concentration of nearly 30 µM ¹⁵N-NO₃⁻ (and an enrichment of nearly 30%, the ¹⁴N-NO₃⁻ concentration being nearly 100 µM). Shortly after the addition of the labelled nitrate solution (t0) and after 2.5, 5, 7.5 and 10 h of dark incubation (t1, t2, t3, t4, respectively), water samples (12 mL) were collected from each liner, filtered, and immediately frozen for later analysis. From a water phase of 800 mL in each microcosm, a total of 12 mL × 5 samplings = 60 mL (<8% of the water volume) were collected during the course of the experiment. All water samples were then treated as previously described in order to quantify ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ concentrations.

2.7.2. Calculation of ¹⁴N-NO₃⁻-Efflux from Sediments

The total flux of ¹⁴N-NO₃⁻ out of the sediment (¹⁴N-NO₃⁻-efflux) was calculated with two methods. The first (A) is based on linear regressions applied to the concentration changes of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ during the course of the experiment, as detailed in the graphical representation of Figure 2, while the second (B) is based on the equation proposed by [17].

Method A

The slope of the regression between measured ¹⁵N-NO₃⁻ concentration and time (S15m) integrates all ¹⁵N-NO₃⁻ consuming processes (i.e., uptake, denitrification, DNRA) (Figure 2). Knowing the ratio between ¹⁵N-NO₃⁻ and ¹⁴N-NO₃⁻ concentrations in the water column at t0 and the slope S15m, the analogous theoretical slope S14t, integrating all the ¹⁴N-NO₃⁻ consuming processes, can be calculated (S14t = S15m × [¹⁴N-NO₃⁻/¹⁵N-NO₃⁻]). If significant, the difference between S14t and the slope of measured ¹⁴N-NO₃⁻ trend (DS14 = S14m − S14t) represents the ¹⁴N-NO₃⁻-efflux (Figure 2). The latter in fact is sustained by nitrification of ¹⁴N-NH₄⁺ within sediments, occurring also for the small sedimentary pool of ¹⁵N-NH₄⁺ (0.366, which is the natural abundance of ¹⁵N in the ammonium pool). The ¹⁴N-NO₃⁻-efflux (µmol N m⁻² h⁻¹) is then calculated according to Equation (6):

$$\left[{}_{}{}^{14}\mathrm{N}-{\mathrm{NO}}_{3 \mathrm{efflux}}^{-}\right]= \left(\mathrm{S}14\mathrm{m}-\mathrm{S}14\mathrm{t}\right) \times 10 \times \mathrm{h} ,$$

(6)

where S14m and S14t (µmol N L⁻¹h⁻¹) are the measured and theoretical slopes of the regression between unlabeled nitrate concentration and time, h is the height of the water phase (cm), and 10 is a unit conversion factor.

Method B

According to [17], the ¹⁴N-NO₃⁻-efflux (µmol N m⁻² h⁻¹) is calculated with Equation (7):

$$\left[{}_{}{}^{14}\mathrm{N}-{\mathrm{NO}}_{3 \mathrm{efflux}}^{-}\right]= \left(\frac{\mathrm{Ci} \times \left(\mathrm{f}-\mathrm{i}\right)}{0.366 -\mathrm{f}}\right) \times 10 \times \frac{\mathrm{h}}{\mathrm{t}} ,$$

(7)

where Ci is the initial concentration of ¹⁵N-NO₃⁻ (µM, t0), f and i are the final (t4) and initial (t0) fractions of ¹⁵N-NO₃⁻ in the total ¹⁴⁺¹⁵N-NO₃⁻ pool, respectively, 0.366 is the natural abundance of ¹⁵N in the ammonium pool, h (cm) is the height of the water phase, and t (h) is the incubation time.

2.7.3. Measurement of Coupled Nitrification–Denitrification

The remaining eight microcosms were incorporated into liners and the isotope pairing technique was applied [5,20]. Briefly, ¹⁵N-NO₃⁻ was added to the water phase to a final concentration of 30 µM, as in previous experiments. Shortly after the addition, a subsample (12 mL) was collected in order to analyze the ¹⁵N-NO₃⁻ to ¹⁴N-NO₃⁻ ratio in the water phase, as previously explained. Then, the cores were immediately closed at the top with a rubber stopper and incubated in the dark for 4 h. At the end of the incubation, all cores were gently slurred and a subsample of the slurry collected, transferred into exetainers, and poisoned with 200 µL of ZnCl₂ 7 M. ²⁹N₂ and ³⁰N₂ produced via denitrification were analyzed by MIMS and calculations performed according to the assumptions and equations reported by [5]. It was then possible to calculate total denitrification and the contribution of the denitrification of nitrate diffusing to anoxic sediments from the water column (D_W) and the denitrification of nitrate produced within sediments via nitrification (D_N). We considered the IPT a reliable method as the anaerobic oxidation of ammonium to molecular nitrogen is generally not detected in eutrophic freshwater sediments [21]. However, we excluded the occurrence of anammox from the Mincio River sediments with a specific assay (see Section 3.5).

2.8. Statistical Analyses

In the nitrification experiment, the values of total N-NO₃⁻-efflux were tested for normality using the Kolmogorov–Smirnov test. As the assumptions for normality were met, two-way analysis of variance (ANOVA) was employed to test for differences between the rates measured in the two treatments and between the methods utilized for the calculations.

3. Results and Discussion

3.1. Slurry Concentration Series

In the range from 42 to 500 mL of sediment per liter of slurry (from 0.5 to 6 mL in a final volume of 12 mL), the recovery of the added ¹⁵N-NO₃⁻ (50 µM) was the same and averaged 89% (data not shown). This confirms previous findings that the enzymatic capacity of the denitrifier community is very large in the sediments we used [15]. When relatively low ¹⁵N-NO₃⁻ concentrations are expected, a very small sediment volume can be added to the water samples to perform quantitatively the reduction to N₂. We standardized the method fixing the slurry concentration at 167 mL L⁻¹ (2 mL of sediment and 10 mL of water sample) as most of our study areas, within the Po River Plain, may display very high NO₃⁻ concentrations, up to 1000 µM [22].

3.2. Time Series Experiment

The concentration of ¹⁵N-NO₃⁻ calculated from the production of ³⁰N₂ during the course of the time series experiment is reported in Figure 3. Concentrations of ¹⁵N-NO₃⁻ (nearly 2, 20, 200 and 1000 µM) in the slurries at t0 are estimated, as they were not rigorously checked by spectrophotometric analyses. Nitrate reduction and the accumulation of labelled gas reached a plateau in <5 h when the ¹⁵N-NO₃⁻ concentrations in the slurries were 2, 20, and 200 µM, while it took longer (nearly 20 h) at 1000 µM (Figure 3). The plateau was set at the concentration of ¹⁵N-NO₃⁻ (± 5%) recovered after 48 h of slurry incubation and the appropriate incubation time was set graphically. During the anoxic incubation, we found trace amounts of ²⁹N₂, due to microbial reduction of ¹⁴N-NO₃⁻ in the K¹⁵NO₃ salt used for this experiment (98% ¹⁵N). Based on the results of this experiment, we set the incubation time of the following experiments at 20 h to optimize production of N₂ regardless of the initial concentration of ¹⁵N-NO₃⁻.

3.3. Concentration Series Experiment

We tested the denitrification capacity of the riverine sediments we employed and the recovery of the added ¹⁵N-NO₃⁻ with a concentration series experiment (Figure 4). With the slurry concentration of 167 mL L⁻¹, we obtained a rather good recovery of the added ¹⁵N-NO₃⁻, averaging 86% over a wide range of labelled nitrate levels (from 0.5 to 800 µM). Thereafter, the enzymatic capacity of the denitrifier population in the sediment was saturated and the production of labelled N₂ became independent from the ¹⁵N-NO₃⁻ concentration. The linear regression reported represents the calibration curve of this method, under the specific experimental settings described in Table 1. The calibration allows for converting the measured labelled dinitrogen into the original ¹⁵N-NO₃⁻ concentration in the water samples. Results from this test support the use of this assay over a wide range of ¹⁵N-NO₃⁻ concentrations. It is however important to remark that each sediment needs to be tested in advance, and a similar calibration should be always performed together with the processing of samples from experiments with unknown ¹⁵N-NO₃⁻ concentrations.

3.4. Quantification of Labelled and Unlabeled N-NO₃⁻

Standard solutions containing a wide range of combinations of labelled and unlabeled N-NO₃⁻ underwent reduction in order to quantify the recovery of ¹⁵N and ¹⁴N. Due to the large background concentration of ²⁸N₂, the ¹⁴N is calculated from the measured ²⁹N₂ and ³⁰N₂ and not readily measured [5]. This may lead to large errors when the atom% of the ¹⁵N in the total N-NO₃⁻ pool is very low and the produced ²⁹N₂ and ³⁰N₂ deviate from the predictions of the binomial distribution. In the range between 16.5 to 83.5 atom% of ¹⁵N, we recovered the same amount of ¹⁵N atom% in the N₂ pool (Figure 5a). Outside this range, calculations produced large errors in the estimate of ¹⁴N from the measured ²⁹N₂ and ³⁰N₂. This suggests, and confirms, that the enrichment of the ¹⁴N-NO₃⁻ pool with labelled nitrate should be >15% to obtain precise measurements and calculations of ¹⁵N and ¹⁴N.

We performed a second validation of this method using standard solutions containing slightly different concentrations of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻, in order to see the accuracy of the method (Table 1 and Figure 5b). This second experiment confirmed the previous one and produced a very good fit between the recovered and expected concentrations of the labelled and unlabeled nitrate forms.

3.5. Anammox Assay

The ratio between the theoretical and measured ²⁹N₂ concentrations along a ¹⁵N-NO₃⁻ concentration series (from 5 to 200 µM) was 1, and it was independent of the labelled nitrate concentration, suggesting that anammox was not contributing to N₂ production in our anoxic slurries (Figure 6) [16]. The theoretical ²⁹N₂ concentration was predicted by the binomial distribution of the ^28/29/30N₂ accumulated during the incubation, assuming that bacterial denitrification was the only N₂ producing process. It is not surprising that anammox is not a relevant process in organic freshwater sediments [20,23,24], but this test should always be performed in order to verify the assumptions of [5] and avoid incorrect estimations of ²⁸N₂ [16].

3.6. Measurements of ¹⁵N-NO₃⁻ and ¹⁴N-NO₃⁻ to Determine Nitrification Rates in Bioturbated Sediments

3.6.1. ¹⁴N-NO₃⁻-Efflux from Sediments

In both treatments C and H, the application of ¹⁵N-NO₃⁻ dilution allowed for calculating a net ¹⁴N-NO₃⁻-efflux, significantly lower, by nearly 30%, in the presence of the oligochaetes (two way ANOVA, p < 0.05, Figure 7a). Results from calculations based on time series measurement of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ (method A) tended to be lower than those based on method B [17] but differences were not significant (two way ANOVA, p > 0.05). Calculations with method B produced very comparable results if the term considering the natural abundance of ¹⁵N-NH₄⁺ (0.366) was excluded from Equation (7) (method B *).

3.6.2. Coupled Nitrification–Denitrification

Denitrification rates of the ¹⁴N-NO₃⁻ produced within sediments via nitrification (D_N, Figure 7b) were not statistically different in bare and bioturbated sediments (t-test, p > 0.05), and averaged 245 ± 30 µmol N m⁻² h⁻¹. Rates of denitrification of water column ¹⁴N-NO₃⁻, not reported, were 102 ± 23 and 119 ± 10 µmol N m⁻² h⁻¹ in C and H, respectively. They tended to be higher in the presence of S. tamesis but differences were not significant (t test, p > 0.05).

Overall, calculated nitrification rates (¹⁴N-NO₃⁻-efflux + D_N) in the riverine sediments used in our experiment were among the highest reported in the literature, averaging 408 ± 40 and 350 ± 31 µmol N m⁻² h⁻¹ in C and H, respectively [25,26,27]. It is surprising, but not entirely novel, that a deep burrower as S. tamesis is not enhancing but depressing rates of nitrification, and it is not stimulating the rates of denitrification. Similar results are reported for another opportunistic polychaete that has invaded the Baltic Sea (Marenzelleria spp.) and is able to cope with anoxic and chemically reduced conditions in the pore water. Marenzelleria was demonstrated to stimulate more the dissimilative reduction of nitrate to ammonium (DNRA) than denitrification, likely due to the occurrence of free sulphides in its burrow, which are toxic to denitrifiers [28]. We speculate that S. tamesis adds to the list of pioneer species able to colonize very organic sediments, not requiring frequent ventilation of their burrows or acting as conveyors (sensu [29]). These bioturbators in fact ventilate their burrows extracting anoxic pore water from surrounding sediments, without creating significant toxic niches and without any stimulation of nitrifiers growth.

4. Conclusions

The proposed approach represents a relatively easy method to quantify accurately dissolved N-NO₃⁻ in mixtures of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ in freshwater. Instead of using pure cultures of denitrifiers, that may require specific skills, or different chemicals to reduce the nitrate pool, we used natural populations of denitrifiers from the sampling site where we performed the nitrification experiment, after calibration of their performances. As we used organic sediments from a nitrate-rich river, our slurry displayed a very large denitrification potential, being able to reduce quantitatively standards up to 800 µM ¹⁵N-NO₃⁻, but no measurable anammox. The recovery of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ from standards with different ratios of the two isotopes was also satisfactory, along a wide range of ¹⁵N atom%. It is important to always check for the concurrence of anammox, which would result in the violation of the assumptions related to the binomial distribution of ^28/29/30N₂ [5]. Our specific test demonstrated that, in the Mincio River sediments, the production of ²⁹N₂ was unaffected by increasing concentrations of ¹⁵N-NO₃⁻, suggesting a negligible role of anammox bacteria in N₂ production. However, in the case of adopting the present approach, the sediments and their performances need to be specifically tested. The accurate measurement of ¹⁴N-NO₃⁻ and ¹⁵N-NO₃⁻ concentrations has different applications in ecological studies of microbial N transformations. One of these is the quantification of ammonium oxidation to nitrate in bioturbated sediments, via the ¹⁵N-NO₃⁻ dilution technique. Nitrification rates in intact sediments are generally modelled and seldom experimentally measured as compared to processes such as denitrification or nitrate ammonification. However, in oligotrophic aquatic ecosystems, nitrification is the only source of nitrate for the above-mentioned processes. The application of our slurry approach to nitrification measurements in bioturbated sediments allowed to demonstrate that not all burrowers, as generally reported in the literature, provide ecosystem services such as the increase of the toxic sediment volume and the stimulation of N loss via coupled nitrification–denitrification.