This method estimates sediment denitrification rates in sediment cores using the acetylene inhibition technique and microsensor measurements of the accumulated N2O. The protocol describes procedures for collecting the cores, calibrating the sensors, performing the acetylene inhibition, measuring the N2O accumulation, and calculating the denitrification rate.
Denitrification is the primary biogeochemical process removing reactive nitrogen from the biosphere. The quantitative evaluation of this process has become particularly relevant for assessing the anthropogenic-altered global nitrogen cycle and the emission of greenhouse gases (i.e., N2O). Several methods are available for measuring denitrification, but none of them are completely satisfactory. Problems with existing methods include their insufficient sensitivity, and the need to modify the substrate levels or alter the physical configuration of the process using disturbed samples.This work describes a method for estimating sediment denitrification rates that combines coring, acetylene inhibition, and microsensor measurements of the accumulated N2O. The main advantages of this method are a low disturbance of the sediment structure and the collection of a continuous record of N2O accumulation; these enable estimates of reliable denitrification rates with minimum values up to 0.4-1 µmol N2O m-2 h-1. The ability to manipulate key factors is an additional advantage for obtaining experimental insights. The protocol describes procedures for collecting the cores, calibrating the sensors, performing the acetylene inhibition, measuring the N2O accumulation, and calculating the denitrification rate. The method is appropriate for estimating denitrification rates in any aquatic system with retrievable sediment cores. If the N2O concentration is above the detection limit of the sensor, the acetylene inhibition step can be omitted to estimate the N2O emission instead of denitrification. We show how to estimate both actual and potential denitrification rates by increasing nitrate availability as well as the temperature dependence of the process. We illustrate the procedure using mountain lake sediments and discuss the advantages and weaknesses of the technique compared to other methods. This method can be modified for particular purposes; for instance, it can be combined with 15N tracers to assess nitrification and denitrification or field in situ measurements of denitrification rates.
Anthropogenic alteration of the nitrogen cycle is one of the most challenging problems for the Earth system1. Human activity has at least doubled the levels of reactive nitrogen available to the biosphere2. However, there remain large uncertainties regarding how the global N cycle is evaluated. A few flux estimates have been quantified with less than ±20% error, and many have uncertainties of ±50% and larger3. These uncertainties indicate the need for accurate estimations of denitrification rates across ecosystems and an understanding of the underlying mechanisms of variation. Denitrification is a microbial activity through which nitrogenous oxides, mainly nitrate and nitrite, are reduced to dinitrogen gasses, N2O and N24. The pathway is highly relevant to the biosphere availability of reactive nitrogen because it is the primary process of removal5. N2O is a greenhouse gas with a warming potential nearly 300 times that of CO2 over 100 years, and it is the current major cause of stratospheric ozone depletion due to the large quantities being emitted6,7.
In the following, we present a protocol for estimating sediment denitrification rates using cores and N2O microsensors experimentally (Figure 1). Denitrification rates are estimated using the acetylene inhibition method8,9 and measurements of the accumulation of N2O during a defined period (Figure 2 and Figure 3). We demonstrate the method by applying it to mountain lake sediments. This case study highlights the performance of the method for detecting relatively low rates with minimal disturbance to the physical structure of the sediments.
Denitrification is particularly difficult to measure10. There are several alternative approaches and methods, each with advantages and disadvantages. Drawbacks to available methods include their use of expensive resources, insufficient sensitivity, and the need to modify the substrate levels or alter the physical configuration of the process using disturbed samples10. An even more fundamental challenge to measuring N2 is its elevated background levels in the environment10. The reduction of N2O to N2 is inhibited by acetylene (C2H2)8,9. Thus, denitrification can be quantified by measuring the accumulated N2O in the presence of C2H2, which is feasible due to low environmental N2O levels.
The use of C2H2 to measure denitrification rates in sediments was developed about 40 years ago11, and the incorporation of N2O sensors occurred about 10 years later12. The most widely applied acetylene-based approach is the "static core". The accumulated N2O is measured during an incubation period of up to 24 h after the C2H2 is added to the headspace of the sealed sediment core10. The method described here follows this procedure with some innovations. We add the C2H2 by bubbling the gas in the water phase of the core for some minutes, and we fill all the headspace with sample water before measuring the accumulation of N2O with a microsensor. We also include a stirring system that prevents the stratification of the water without resuspending the sediment. The procedure quantifies the denitrification rate per sediment surface area (e.g., µmol N2O m-2 h-1).
The high spatial and temporal variation of denitrification presents another difficulty in its accurate quantification10. Usually, N2O accumulation is measured sequentially by gas chromatography of headspace samples that are collected during the incubation. The method described provides improved monitoring of the temporal variation of the N2O accumulation, because the microsensor provides a continuous signal. The microsensor multimeter is a digital microsensor amplifier (picoammeter) that interfaces with the sensor(s) and the computer (Figure 1a). The multimeter allows several N2O microsensors to be used at the same time. For instance, up to four sediment cores from the same study site can be measured simultaneously to account for the spatial variability.
The core approach barely disturbs the sediment structure compared to some other methods (e.g., slurries). If the integrity of the sediments is altered, this leads to unrealistic denitrification rates13 that are only adequate for relative comparisons. Higher rates are always obtained with slurry methods compared to core methods14, because the latter preserves the limitation of denitrification by substrate diffusion15. Slurry measures cannot be considered representative of in situ rates16; they provide relative measures for comparisons made with the exact same procedure.
The method described is appropriate for estimating denitrification rates in any sediment type that can be cored. We particularly recommend the method for performing experimental manipulations of some of the driving factors. Examples are experiments that modify nitrate availability and temperature as needed for estimating the energy activation (Ea) of denitrification17 (Figure 2).
Figure 1: Experimental setup. (a) General experimental setup to estimate sediment denitrification rates using cores and N2O microsensors. The incubation chamber ensures darkness and controlled-temperature (±0.3 °C) conditions. Five intact sediment cores can be processed simultaneously using their respective N2O sensors. (b) N2O sensor calibration chamber. We adapted it with rubber stoppers and syringes to mix the N2O water (see protocol step 3.4.3). There is a thermometer to control the water temperature. (c) Close-up of a sediment core sample with the sensor inserted into the central hole of the PVC cover and the joints sealed with adhesive tape. The stirrer is hanging in the water, and the electromagnet is close to it and fixed to the external part of the acrylic tube. (d) Close-up of the N2O microsensor tip protected by a metal piece. (e) A sediment core that has just been recovered. It was sampled from a boat in a deep lake; the acrylic tube with the core is still fixed to the messenger-adapted gravity corer19. See the Table of Materials for all the items needed to perform this method. Please click here to view a larger version of this figure.
1. Preparation
NOTE: Begin this on the day before the measurements are taken.
2. Sediment Core Collection
3. Calibration of the Nitrous Oxide (N2O) Microsensors
4. Core Preparation and Acetylene Inhibition
5. Denitrification (N2O accumulation measure)
6. Final Measurement Steps
7. Denitrification Rate Calculations
A total of 468 denitrification rates were estimated using the protocol above in sediments from Pyrenean mountain lakes over the period 2013-2014. We show some of these results to illustrate the procedure (Figure 2 and Figure 3). In general, the linear model between the N2O concentration and time has good correlation (R2 ≥ 0.9). The slope of the relationship provides an estimate of the denitrification rate (step 7.3; e.g., Figure 2d). If the denitrification activity is very low, the sensor's electronic noise becomes more important and the goodness of fit declines (e.g., sensors 4 and 5 in Figure 2b and Figure 3a). Although the baseline detection limit of N2O is ~0.1 µM in water22, which is an intermediate value concerning alternative methods23, the possibility of accumulating thousands of continuous measurements to filter the noise permits estimates at relatively low denitrification rates, up to ~1 µmol N2O m-2 h-1 (Figure 2 and Figure 3). Lower rates (i.e., ~0.4 µmol N2O m-2 h-1) can be estimated by narrowing the water phase of the core sample to a height of 8 cm (see protocol step 4.2).
Figure 2: Denitrification rate calculations in a temperature dependence experiment. Actual (a and b) and potential denitrification measurements (c–f) are shown. When the temperature of the measurement is decreased (c), at first the sample cools and the sensor signal, which is temperature dependent, declines. (a) A similar event occurs at the start of the incubation in the actual denitrification measurement; the warmer laboratory environment with respect to the incubation conditions produces a cooling of the sample, again accompanied by a decline in the sensor signal. (e) When the temperature is increased, at first the samples warm and the sensor signal increases exponentially instead of linearly. When the samples reach a constant temperature, the sensor signal increases linearly as usual. In all cases, it is possible to calculate the denitrification rates just by using the period of linear N2O accumulation (b, d, and f). (b) Inactive sample 3 is not shown. Please click here to view a larger version of this figure.
Figure 3: Examples of denitrification rate calculations. Actual (a) and potential (b and c) denitrification rates were estimated. We only used the time range with a linear N2O accumulation to calculate the denitrification rate (slope of the linear model). However, in (a), for educational purposes, we show all the measurements (models) with more and less success; we would discard sample 3 due to the high instability of the sensor and sample 2 due to saturation in the N2O accumulation. (a) Samples 4 and 5 with rates of 0.5 and 0.7 µmol N2O m-2 h-1, respectively, are cases of measurements near the detection limit of the method. Please click here to view a larger version of this figure.
The main advantages of the described method are the use of minimally disturbed sediment core samples and the continuous recording of the N2O accumulation. These allow estimation of relatively low denitrification rates that are likely similar to those occurring in situ. Nonetheless, some aspects concerning the coring, sensor performance, and potential improvements are discussed.
An apparently simple but critical step of the method is good core recovery. The sediment/water interface must satisfy three criteria: (i) no modification in its chemical or constituent composition, (ii) no alteration in the water content or void ratio, and (iii) no structure pertubation24. The fewer disturbances suffered by the sample during the entire protocol, the more realistic and closer to in situ conditions will the measured denitrification rate be. There are several devices/techniques for the sediment core collection25, and their selection depends on the water depth. We use a messenger-adapted gravity corer19 for deep samples (Figure 1e) because it is a reasonably light-weight device and can rapidly recover short cores25 (a core sediment of ≥10 cm length is more than enough to encompass the oxic and denitrifying layers in the sediments26,27,28). In coring jargon, "feel" is often referred to as the ability to know the location of the corer (whether it is still in the water column or already in the sediment) and whether it is open or closed25. For intermediate water depths (5-50 m), usually there are no difficulties with feeling. A loss of feeling occurs in deeper water (>50 m) because the movements of the water column may mask the location of the corer25. Feeling may also be lost in shallow water (<3 m) due to lateral drift and wave action25; this is why we use a different method in shallow water, either direct manual coring by scuba diving or dressing in a wader. With this system, the person performing the sampling can see the sediment and choose the exact place before coring; this allows, e.g., the sampling of a sediment core that contains a macrophyte. After sampling, the researcher must continue to work carefully to minimally disturb the sediment core sample during the rest of the protocol, especially when performing acetylene inhibition by bubbling.
Some details must be considered when using N2O microsensors. The sensor software provides a continuous visualization (strip chart) of the sensor signal (background frequency of 1000 Hz)29. These raw data and the strip chart (e.g., Figure 2a) can be saved. It is necessary to check the correct behavior of the sensor after its polarization (e.g., when returning from field collection before step 4). In particular, a low (<20 mV) and constant base signal is expected when it is submerged in N2O-free water. Recalibrate the sensor shortly (~2 h) after starting its use; if it has already been used for some days, the interval can be extended (~24 h)18. To minimize recalibrations, keep the sensor polarized unless it is not used for several days18. Over time, a change in the sensor signal may occur, up to 50% in months, which is due to a different permeability of its membrane18. The lower the electronic interference in the laboratory, the more constant and stable will be the sensor signal. In that sense, using a UPS improves the quality of the electrical energy that reaches the measurement device by filtering the voltage fluctuations. The sampling interval, selected in the Logger tab, is different from the background frequency. Each registered point is generated from the average of many measurements. The sampling interval (up to 10 s) indicates the frequency with which a data point is recorded. The number of measurements per unit of time used in the average is defined by the background frequency29. For instance, if we set a sampling frequency of 5 s and a background frequency of 500 measurements per second, then the data points are recorded every 5 s and the average of the 500 samples per second is measured during the previous 5 s. We record the sensor signal every 5 min (sampling interval) and set the background frequency to 1000 measurements per second. The study system must be known to select the correct sampling interval without "averaging" expected fluctuations. In highly active systems, short sampling intervals are recommended, while longer intervals allow optimizing the computer's memory29. Some possible interfering substances (H2S, NO, and CO2) can affect the N2O sensor's signal22. The sensor is calibrated with deionized water, but the samples can contain interfering substances and modify the sensor's reference signal. This situation could explain why negative values appear in samples 2 and 5 in Figure 2b and Figure 3a, respectively. However, when the objective is to estimate the denitrification rate, the exact level of N2O is not the key parameter. What is key is the slope of the linear model (evidencing a linear accumulation of N2O). Finally, it is necessary to work with a fixed temperature because the response of the N2O sensor changes with temperature (Figure 4).
Simple modifications or additions to the protocol also enable (i) characterization of the environmental conditions controlling the measured denitrification rates, (ii) estimation of the potential denitrification rates by simulating the response to a driving gradient (e.g., nitrate), and (iii) estimation of the sediment N2O emission rates by skipping the C2H2 inhibition Depending on the study aims, several complementary measurements can be made: (i) just after recovering the core, in situ conditions, e.g., temperature; (ii) before the measurement, samples of the water phase, e.g., [NO3–]; and (iii) after the measurement, extrusions and slices of the core at different resolutions (mm-cm)25,30, following the procedures explained by P. T. Schwing et al.30.
To measure the potential denitrification rates, add nitrate to the water-phase of the core (e.g., Figure 2 and Figure 3) as described in C. Palacin-Lizarbe, L. Camarero and J. Catalan17. If doing so, add the nitrate before the C2H2 inhibition (step 4.3). Also, if nitrate is added, it is advisable to also add carbon (C; e.g., acetate) and phosphorus (P) to maintain the in situ stoichiometric proportions of C, N, and P (e.g., in the surface sediment). This will prevent the limitation of denitrification by these elements31,32, and will also keep the C/N ratio that can influence the dominance of the nitrate consumption process (i.e., denitrification versus dissimilatory nitrate reduction to ammonium (DNRA))4. Anoxia can be fixed by bubbling an N2-CO2 mixture for a few minutes after the nitrate addition, to prevent oxygen interference with denitrification; however, note that this leads to a blockage of nitrification. To calculate sediment N2O emission rates, omit the C2H2 inhibition (step 4.3). However, keep in mind that, as far as it is currently known in aquatic ecosystems, N2O emissions are proportionally low compared to N2 emissions (0%-4.3%)33, so it is possible that the accumulated N2O will be below the detection limit. If this is the case, an option is to add nitrate to increase the emitted N2O, calculating potential N2O emissions.
The main weakness of the method is the inhibition of nitrification by C2H210,34. During the incubation, this inhibition of nitrification and the incomplete inhibition of N2O reduction may become apparent, as both are very time dependent. For instance, the starting N2O accumulation rate must reveal the real denitrification rate and progressively decay as the nitrate availability drops and N2O diffuses into the nitrate free zone, where it is reduced35. Therefore, an estimated denitrification rate can be considered valid only if the readings show a linear accumulation of N2O10.
The method described estimates a denitrification rate per area that integrates the entire sediment activity. In this respect, there is some uncertainty about the radius of action of the acetylene inhibition when bubbling the gas in the aqueous phase of the sample. It is assumed that, at least, inhibition of the surficial layer of the sediment occurs, which is the one with the highest denitrification rates26,27.
Possible improvements to this method are its combined use with 15N tracers and modifications that could allow the measurement of denitrification in situ. 15N tracer methods can be used to determine the proportion of nitrification-denitrification coupling occurring in the samples36, and it can also account for other N flux processes besides denitrification (e.g., anammox and dissimilatory nitrate reduction to ammonium (DNRA))13,37. However, these methods have the drawback of changing the substrate concentration10. A. Behrendt, D. de Beer and P. Stief 26 use a method combining N2O microsensors, C2H2 inhibition, and 15N tracers to analyze the vertical activity distribution of dissimilatory nitrate reduction processes (denitrification and DNRA) in sediments. They made vertical profiles in the sediment by penetrating the sediment with the sensors. The main difficulty in measuring denitrification in situ is the ability to handle a nonconstant temperature environment. It is necessary to record the N2O accumulation and temperature simultaneously and then correct the N2O sensor's signal by the temperature dependence during the denitrification rate calculations. This correction requires a previous analysis of the temperature dependence of the N2O signal for each sensor. The sensors are handmade, and each one responds differently to temperature (e.g., sensor 1 shows a higher temperature dependence than the others in Figure 2c, e).
Figure 4: Temperature dependence of the N2O microsensor response. The different slopes of the linear model of the sensor signal versus the temperature at each N2O concentration shows the temperature effect on the sensor's signal. Please click here to view a larger version of this figure.
The authors have nothing to disclose.
The Spanish Government provided funds through the Ministerio de Educación as a predoctoral fellowship to C.P-L. (FPU12-00644) and research grants of the Ministerio de Economia y Competitividad: NitroPir (CGL2010-19737), Lacus (CGL2013-45348-P), Transfer (CGL2016-80124-C2-1-P). The REPLIM project (INRE – INTERREG Programme. EUUN – European Union. EFA056/15) supported the final writing of the protocol.
Messenger-adapted gravity corer | – | – | Reference in the manuscript. Made by Glew, J. |
Sampling tube | – | – | Acrylic. Dimensions: 100 cm (h) × 6.35 cm (d) × 6.50 cm (D). Sharpen the edge of the sampling tube that penetrates into the sediment to minimize the disturbance in the recovered sediment core sample. |
Handheld sounder | Plastimo | 38074 | Echotest II Depth Sounder. |
Rubber stopper | VWR | DENE1012114 | With two holes, used to mix the N2O-water in the calibration chamber. Dimensions: 20 mm (h) × 14 mm (d) × 18 mm (D) (3 mm hole (D)). |
Rubber stopper | VWR | 217-0125 | To seal the bottom part of the methacrylate tube and to sample in shallow water bodies. Dimensions: 45 mm (h) × 56 mm (d) × 65 mm (D). |
PVC cover | – | – | To seal the top side part of the acrylic tube. Dimensions: 45 mm (h) × 56 mm (d) × 65 mm (D). Dimensions: 65 mm (D). |
Adhesive tape | – | – | Waterproof. To ensure all joints (PVC cover sampling tube and PVC cover sensor) and to avoid water leaks. |
Thermometer | – | – | Portable and waterproof, to measure the temperature in the water overlying the sediment just after sampling the cores. |
GPS | – | – | To save the location of a new sampling site or to arrive at a previous site. |
Wader | – | – | For littoral or shallow site samplings. |
Boat | – | – | An inflatable boat is the best option for its lightness if the sampling site is not accessible by car. |
Rope | – | – | Rope with marks showing its length (e.g., marked with a color code to distinguish each meter). |
N2O gas bottle and pressure reducer | Abelló Linde | 32768-100 | Gas bottle reference. |
C2H2 gas bottle and pressure reducer | Abelló Linde | 32468-100 | Gas bottle reference. |
Tube used to evacuate the excess of water | – | – | Consists of a solid part (e.g., a 5 ml pipette tip without its narrowest end) inserted in a silicone tube. |
Nitrous Oxide Minisensor w/ Cap | Unisense | N2O-R | We use 4 sensors at a time. |
Microsensor multimeter 4 Ch. 4 pA channels | Unisense | Multimeter | Picoammeter logged to a laptop. The standard device allows for 2 sensor picoammeter connections (e.g., N2O sensor), one pH/mV and a thermometer. We ordered a device with four picoammeter connections, allowing the use of 4 N2O sensors simultaneously. |
SensorTrace Basic 3.0 Windows software | Unisense | Sensor data acquisition software. | |
Calibration Chamber incl. pump | Unisense | CAL300 | Calibration chamber. We tuned it with rubber stoppers and syringes to mix the N2O-water without making bubbles. |
Incubation chamber | Ibercex | E-600-BV | Indispensable equipment for working at a constant temperature (±0.3 °C). It also allows control of the photoperiod. |
Electric stirrer | – | – | Part of the stirring system. It hangs in the water, overlying the sediment subject, by a fishing line that is hooked to the PVC cover. |
Electromagnet | – | – | Part of the stirring system. It is fixed to the outside of the acrylic tube, approximately at the same level as the stirrer. It is activated episodically (ca. 1 on-off per s) by a circuit, attracting the stirrer when it is on and releasing it when it is off, thereby generating the movement that agitates the water. |
Electromagnetic pulse circuit | – | – | Part of the stirring system. It is connected by wires to the electromagnet and sends pulses of current that turn the electromagnet on and off. |
Uninterruptible power supply (UPS) | – | – | It improves the quality of the electrical energy that reaches the measurement device, filtering the highs and low of the voltage, thereby ensuring a more constant and stable N2O sensor signal. |