SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (41)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Laughlin, R. J.
Right arrow Articles by Stevens, R. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Laughlin, R. J.
Right arrow Articles by Stevens, R. J.
Agricola
Right arrow Articles by Laughlin, R. J.
Right arrow Articles by Stevens, R. J.
Related Collections
Right arrow Isotopes
Right arrow Nitrogen
Right arrow Soil Biochemistry
Soil Science Society of America Journal 66:1540-1548 (2002)
© 2002 Soil Science Society of America

DIVISION S-3—SOIL BIOLOGY & BIOCHEMISTRY

Evidence for Fungal Dominance of Denitrification and Codenitrification in a Grassland Soil

Ronald J. Laughlin and R. James Stevens*

Dep. of Agriculture and Rural Development, Agricultural and Environmental Science Division, Newforge Lane, Belfast BT9 5PX, UK

* Corresponding author (jim.stevens{at}dardni.gov.uk)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 Substrate Induced Respiration...
 Headspace Sampling and Soil...
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Fungi are capable of nitrification and denitrification and often dominate the microbial biomass of temperate grassland soils. We determined the contributions of bacteria and fungi to N2O and N2 production in a grassland soil from Northern Ireland by combining the substrate-induced respiration inhibition method and the 15N gas-flux method. Streptomycin (C21H39N7O12) was used as the bacterial inhibitor and cycloheximide (C15H23NO4) as the fungal inhibitor. By labeling the NH4 and NO3 pools, we tested the hypothesis that fungi produce N2O and N2 solely by the reduction of NO3. Cycloheximide decreased the flux of N2O by 89% and streptomycin decreased the flux by 23%, indicating that fungi were responsible for most of the N2O production. All of the N2O was derived from NO3 reduction. Labeled N2 was only detected in control and streptomycin treatments. The distribution of the 15N atoms in the labeled N2 indicated that the source of the labeling was predominantly the NO3 pool, but that the process of formation was not dominated by denitrification. Codenitrification, where a 15N atom from labeled nitrogen dioxide (NO2) combines with a 14N atom from a natural abundance source, was proposed as the process forming labeled N2. About 92% of the labeled N2 was estimated to be due to codenitrification and 8% due to denitrification. The flux of N2O was always greater than the flux of N2, the mole fraction of N2O averaging 0.7. Fungal denitrification could be of ecological significance because N2O is the dominant gaseous end product.

Abbreviations: IRMS, isotope-ratio mass spectrometry • NO2, nitrogen dioxide • SIRIN, substrate-induced respiration inhibition


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 Substrate Induced Respiration...
 Headspace Sampling and Soil...
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
DENITRIFICATION, defined as the dissimilatory reduction of NO3 or NO2 to N2O and N2, has been considered a prokaryotic process for more than a century and has been extensively studied in several bacteria (Zumft, 1997). By the 1970s, it became clear that denitrification is a function of eukaryotes as well as bacteria. Yeasts (Tsuruta et al., 1998) and filamentous fungi (Bollag and Tung, 1972; Shoun et al., 1992) have been shown to be capable of denitrification. Many fungi lack N2O-reductase so N2O is the major product of fungal denitrification (Shoun et al., 1992). Fungi can use NO3 as an alternate electron acceptor to O2 for respiration and can perform aerobic respiration and denitrification simultaneously (Zhou et al., 2001). The occurrence of fungal denitrification in soil could be of ecological significance if N2O is the dominant gaseous end product, since N2O is a radiatively active trace gas and N2 is not. A few species of fungi can produce N2 by a codenitrifcation process where one N atom from NO2 combines with one from a source other than NO2 (Tanimoto et al., 1992).

The microbial biomass of temperate soils is often dominated by fungi (Ruzicka et al., 2000) and the proportion of fungi in the biomass can be affected by agricultural systems and land use. A less intensive no-tillage system as opposed to a more intensive conventional tillage system can favour fungi (Frey et al., 1999). The potential for fungi to produce N2O has been shown in two studies using woodland soils (Castaldi and Smith, 1998; Laverman et al., 2000).

The substrate-induced respiration inhibition (SIRIN) method developed by Anderson and Domsch (1973)(1975) has been widely used to measure the fungal and bacterial biomass in soil. When glucose was added to soil samples, Anderson and Domsch (1973) observed that respiration was increased to a constant level for 2 to 8 h before the liberation of CO2 increased because of proliferation of the soil microorganisms. The new respiration level, which persisted during the lag phase of growth of the microbial population, was called maximum initial response and was indicative of the size of the native biomass. By adding a bacterial inhibitor (streptomycin) and a fungal inhibitor (cycloheximide), Anderson and Domsch (1975) developed a method to measure the relative bacterial and fungal contributions to the respiring soil biomass. The SIRIN method and direct microscopy gave similar results for the proportions of bacteria and fungi in two grassland soils (Lin and Brookes, 1999a).

In this laboratory study, we determined the relative contributions of bacteria and fungi to denitrification in a grassland soil by combining the SIRIN method and the 15N gas-flux method (Mosier and Schimel, 1993). By labeling the NH4 and NO3 pools, we tested the hypothesis that fungi produce N2O and N2 solely by the reduction of NO3 and not by oxidation of NH4.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
Substrate Induced Respiration...
 Headspace Sampling and Soil...
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil Characteristics
An acid brown earth (Typic Dystrochrept) with a pH of 6.3 and containing 48% sand, 31% silt, 20% clay, 0.40% total N, and 4.95% total C was collected (0–10 cm deep) from a site at the Agricultural Research Institute of Northern Ireland, Hillsborough, County Down in June 1999. The sward was dominated by perennial ryegrass (Lolium L.) and had been under a three-cut silage regime receiving 300 kg N ha-1 yr-1 for the previous 7 yr. The soil was sieved (<6 mm) and stored at 4°C for 1 wk before use.


   Substrate Induced Respiration-Inhibition Method
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 Substrate Induced Respiration...
Headspace Sampling and Soil...
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The SIRIN procedure was based on the method of Anderson and Domsch (1975), as modified by Beare et al. (1990) for plant residues. Optimal soil moisture content, glucose concentration, and inhibitor concentrations were determined in preliminary experiments according to the criteria of Anderson and Domsch (1975). Soil moisture contents (40, 45, 50, 55, 60, 65, 70, 75, and 80% water-filled pore space), concentrations of glucose (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, and 15 mg g-1 soil), streptomycin sulfate (0, 0.5, 1.0, 1.5, 2.0, 4.0, 6.0, 8.0, 10.0, and 12.0 mg g-1 soil), and cycloheximide (0, 5, 8, 10, 12, 15, and 20 mg g-1 soil) were evaluated. To comply with the criteria of selective inhibition, the additivity ratios for pairs of inhibitor concentrations were determined. Optimum joint concentrations of inhibitors are indicated when the additivity ratio [(A - B) + (A - C)]/(A - D) most nearly approaches 1; where A is CO2 evolved from glucose only; B is CO2 evolved from glucose + streptomycin; C is CO2 evolved from glucose + cycloheximide; and D is CO2 evolved from glucose + streptomycin + cycloheximide. The optimal conditions for the SIRIN method in this soil were 65% water-filled pore space, 5 mg g-1 glucose, 15 mg g-1 cycloheximide, and 3 mg g-1 streptomycin. These conditions were used in the laboratory experiment described here to test the effect of cycloheximide and streptomycin on N2O, N2, and CO2 production after four incubation times from differentially labeled NH4NO3.

Treatments
Sieved soil was subsampled (120 g on an oven-dry basis) into 108 acid-washed, 500-mL Mason jars. The jars were covered with Parafilm (American National Can, Greenwich, CT) to prevent moisture loss but allow gaseous exchange, and left for 16 h at 4°C. Inhibitors were then added in 20 mL of aqueous solution and mixed into the soil to supply streptomycin at 3 mg g-1 oven-dry soil and cycloheximide at 15 mg g-1 oven-dry soil. Distilled water (20 mL) was added to the control treatment. The jars were again covered with Parafilm and left for a further 16 h at 4°C. Glucose and labeled NH4NO3 (NH154NO3, 15NH4NO3, or 15NH154NO3) were added together to each jar in 5 mL of aqueous solution, to supply glucose at 5 mg g-1 oven-dry soil and NH4-N and NO3-N at 6.74 µmol N g-1 oven-dry soil. The labeled N moieties of the N treatments were enriched to 60 atom% excess in 15N. The bulk density of the soil was 0.95 g cm-3 and an additional 4 mL of distilled water was added to each jar to attain a water-filled pore space of 65%. This moisture content (0.38 g H2O g-1 oven-dry soil) had previously been determined as the optimal value for maximum respiration. There were three replicates of each treatment arranged randomly in an incubator at 22°C. At 3, 6, 10, and 24 h after the addition of glucose, we measured the fluxes of N2O, N2, and CO2, the enrichment of the N2O, and the size and enrichment of the mineral N pools.


   Headspace Sampling and Soil Extraction
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 Substrate Induced Respiration...
 Headspace Sampling and Soil...
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
For a 2-h period centered on each of the four sampling times, lids with gas-sampling ports were fitted to 27 jars using an O-ring to form a gas-tight seal. At the end of each 2-h period, the headspace of each jar was sampled twice using a gas-tight syringe fitted with a push button valve. A 12-mL sample was transferred to an evacuated (<100 Pa) septum-capped vial for analysis of 15N in N2 and N2O by isotope-ratio mass spectrometry (IRMS). A 5-mL gas sample was transferred to a He-filled 10-mL septum-capped vial for analysis of CO2 by gas chromatography.

Within 30 min after gas sampling, all of the soil in each jar was extracted by the blending procedure of Stevens and Laughlin (1995). Soil was transferred to a food homogenizer with 200 mL of 3 M KCl. Ten millilters of 2 M KOH were added, and the mixture blended for 30 s. The volume of KOH was sufficient to increase the pH of the mixture to 8.0 and had been determined previously. The soil extraction was performed at pH 8 to prevent the reaction of NO2 with organic-N compounds (Stevens and Laughlin, 1995). A 200-mL portion of each suspension was centrifuged immediately at 2000 x g for 5 min and the supernatant filtered sequentially through GF/D and GF/F glass-fibre papers (Whatman International Ltd, Kent, UK). Filtrates were stored at 4°C prior to analysis within 2 d for concentration and 15N contents of NH4, NO2, and NO3.

Gas Analysis
The concentration and 15N content of N2O and the 15N content of the N2 in each 12-mL vial was determined by automated IRMS as described by Stevens et al. (1993) using a Europa Scientific 20-20 Stable Isotope Analyser interfaced to a Europa Scientific Trace Gas Preparation System ANCA-TG (Crewe, UK) with Gilson autosampler (Anachem, Luton, UK). The ion currents (I) at m/z 44, 45, and 46 enabled concentrations and molecular ratios 45R (45I/44I) and 46R (46I/44I) to be calculated for N2O. The sources of N2O were then apportioned into the fraction derived from the denitrifying pool of enrichment aD and the fraction d'N = derived from the pool or pools at natural abundance (Arah, 1997). For N2, the ion currents at m/z 28, 29, and 30 enabled molecular ratios 29R (29I/28I) and 30R (30I/28I) to be determined. Differences between the molecular ratios in enriched and normal atmospheres were calculated as {Delta}29R and {Delta}30R. The flux of N2 was calculated by three different methods: (1) using data for {Delta}29R and {Delta}30R to calculate the enrichment of the denitrifying pool (15XN) and then the N2 flux according to Mulvaney and Boast (1986), (2) using {Delta}30R data only and the equation of Mulvaney (1984) assuming that the enrichment of the denitrifying pool was aD (Stevens and Laughlin, 2001a), and (3) using data for {Delta}29R and {Delta}30R to calculate a separate contribution because of codenitrification in addition to N2 from denitrification calculated by Method 2.

Denitrification contributes to {Delta}29R and {Delta}30R whereas codenitrification contributes mostly to {Delta}29R, the {Delta}29R/{Delta}30R ratio always being 272 (Clough et al., 2001). All of {Delta}30R was assumed to be due to denitrification, so {Delta}30R was used to calculate the flux of N2 due to denitrification by Method 2. Using the backsolver facility in Microsoft Excel (Microsoft, Redmond, WA), the value of {Delta}29R that could be attributed to denitrification was then obtained. The difference between the {Delta}29R due to denitrification and the total measured {Delta}29R was assigned to codenitrification. Equation [5] of Clough et al. (2001) for calculating {Delta}29R was rearranged to calculate the fraction of the total moles of N2 in the headspace from codenitrification (dCD):

[1]
where p1 (0.9963) and q1 (0.0037) are atom fractions of 14N and 15N in the natural abundance pool; p2 (0.46) and q2 (0.54) are the atom fractions of 14N and 15N in the enriched NO3 pool from which codenitrification is assumed to occur.

Denitrification rates were expressed as nanomoles of N per gram of oven-dry soil per hour and were corrected for N2O dissolved in the aqueous phase of the soil by using the Bunsen coefficient (Tiedje, 1994). The concentration of CO2 was determined in each 10-mL vial using a Varian Genesis headspace autosampler interfaced to a Varian model 3800 gas chromatograph (Walton-on-Thames, UK) fitted with a 5 m by 2 mm Porapak QS column (80–100 mesh) and a thermal conductivity detector.

Soil Analyses
Soil properties were determined by standard methods (Ministry of Agriculture, Fisheries and Food, 1986). Total N and C were determined using a Carlo Erba N1500 elemental analyzer (Carlo Erba, Milan, Italy). Concentrations of NO3 and NH4 in the KCl extracts were determined by segmented-flow analysis (Technicon Random Access Automated Chemistry System 800+, Bran and Luebbe, Norderstedt, Germany). Nitrate was determined by the sulphanilamide-naphthylethylenediamine method after Cd reduction to NO2. Ammonium was determined by the indophenol blue method (Keeney and Nelson, 1982). Nitrite was determined by a manual colorimetric method based on the sulphanilamide naphthylethylenediamine procedure (Keeney and Nelson, 1982). The 15N contents of the NO3, NO2, and NH4 were determined by methods based on the generation of N2O for IRMS. The production of N2O from NO2 and NO3 is based on the reaction between NO2 and NH2OH in acid conditions, the NO3 having been reduced to NO2 with Cd (Stevens and Laughlin, 1994). The production of N2O from NH4 consists of a diffusion stage where NH3 is absorbed into H2SO4 and an oxidation step where recovered (NH4)2SO4 is oxidized to N2 by alkaline NaOBr, during which N2O is produced as a by-product (Laughlin et al., 1997). For samples containing cycloheximide, the diffusion was performed at pH 8.0 in phosphate buffer instead of MgO for 3 d to prevent the hydrolysis of cycloheximide (Kornfeld and Jones, 1948) producing NH3 at natural abundance and diluting the enrichment of the NH4 pool.

Effect of Inhibitors on Ergosterol in Soil
We checked on the efficiency of SIRIN by assaying for ergosterol, a sensitive and reliable indicator of live fungal biomass (Zelles and Alef, 1995). Jars were set up as described previously using NH4NO3 at natural abundance to provide soil (three replicates) for ergosterol determination after incubation times of 3, 10, and 24 h. The ergosterol content of the soil was determined according to the method of Bentham et al. (1992). Briefly, ergosterol was extracted from two 10 g portions of fresh soil from each jar, one of which was spiked with 100 µg of ergosterol to check on the efficiency of the extraction procedure. Each soil portion was refluxed for 1.5 h in 50 mL of a methanol/ethanol (4:1 v/v) mix to which 4 g of KOH had been added. After cooling, 20 mL of distilled H2O was added to each tube and the extracts recovered by vacuum filtration. The ergosterol was then recovered by solvent extraction with 50 followed by 70 mL of hexane, evaporated to dryness under a stream of N2 and redissolved in 2 mL of hexane/propan-2-ol (98:2, v/v). Analysis was performed with normal-phase HPLC using a Hewlett Packard model 1090 system (Winnersh, Wokingham, UK). A 20-µL aliquot was injected into a 150-mm (4.6 mm i.d.) Lichrosorb Si 60 (10 µm) column (Phenomenx, Macclesfield, UK) eluted with hexane/propan-2-ol (98:2, v/v) at 1.5 mL min-1. Absorbance was measured at 282 nm using a Hewlett Packard Programmmable Fluorescence Detector model 1046A (Hewlett Packard, Winnersh, Wokingham, UK). Results were expressed as micrograms ergosterol per gram of oven-dry soil. The recovery of ergosterol was calculated as

where Ergspiked is the concentration of ergosterol measured in the soil sample spiked with 100 µg of ergosterol, Ergsoil is the ergosterol concentration measured in the soil sample, and Ergadded is the ergosterol concentration due to the added spike.

Calculation of Gross Nitrogen Transformation Rates
The gross N transformation rates for mineralization of NH4, consumption of NH4, nitrification, and consumption of NO3 were calculated for the time interval of zero to 24 h according to the equations of Kirkham and Bartholomew (1954).

Glucose Determination
The determination of the concentration of glucose in KCl extracts of the soils was according to a modification of the anthrone method of Morris (1948). A 5-mL aliquot of the KCl extract was pipetted into a 100-mL digestion tube and cooled in a beaker of ice and water for 10 min. Then 10 mL of anthrone reagent (2 g L-1 17 M H2SO4) was added, keeping the temperature below 25°C. The digestion tube was then placed in a boiling water bath for 5 min and cooled to room temperature. The absorbance of the blue-green complex formed was measured at 625 nm within 40 min of the color development against glucose standards.

Statistical Analyses
Analysis of variance was conducted using Genstat (1993) to determine the significance of treatments on the fluxes of N2O, N2, and CO2, on the values of d'D, aD, and 15XN, on the size and enrichment of the NH4, NO2, and NO3 pools, and on the ergosterol content and recovery in soil. When effects were significant at P < 0.05, the least significant difference (LSD) between means was calculated for P = 0.05.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 Substrate Induced Respiration...
 Headspace Sampling and Soil...
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Confirmation of the Efficacy of the Inhibitors
Evidence for the efficacy of the inhibitors was sought from measurements of respiration. The flux of CO2 into the headspace was unaffected by 15N-labeling, so mean values averaged over all N treatments are presented in Fig. 1 . Cycloheximide significantly (P < 0.05) decreased respiration compared with the control (no inhibitor). Streptomycin had a small inhibitory effect, respiration rates being just significantly less than the control at 6 and 24 h. At the end of the incubation, the glucose concentration in the control and streptomycin treatment had decreased by 60%, and the concentration in the cycloheximide treatment had decreased by 20% (data not shown). For all treatments, respiration remained at constant levels until the 10-h measurement period. A rapid increase in CO2 liberation after 10 h of incubation occurred in the control and streptomycin treatments but in the cycloheximide treatment respiration rate remained at a low, constant level. The lag phase of microbial growth in this study was therefore from 3 to 10 h during which the average inhibition of respiration was 58% due to cycloheximide and 7% due to streptomycin (Fig. 1). These levels of inhibition are similar to those found in other studies (Anderson and Domsch, 1975; Johnson et al., 1996; Lin and Brookes, 1999b). The percentage of fungal biomass as calculated from the respiratory inhibitions observed was 89%, indicating that fungi dominated this soil. In a review of studies in which the proportion of fungi and bacteria had been measured, Ruzicka et al. (2000) concluded that fungi often dominate temperate soils.



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 1. Effect of inhibitors on glucose-induced microbial respiration averaged across ammonium nitrate labeling treatments. (LSD value is for comparing any two means at P = 0.05.)

 
The sum of the inhibitions of respiration by streptomycin and cycloheximide was 65%, so 35% of the respiration was due to fungi and bacteria that existed in the uninhibited biomass. It was assumed that the fungal to bacterial ratio in the uninhibited biomass was the same as indicated by the results of the SIRIN method. This assumption has been tested in other studies by using the SIRIN method in conjunction with other independent methods. Lin and Brookes (1999a) conducted a rigorous test of the SIRIN method for determining the proportions of fungal and bacterial biomass in two UK grassland soils by comparing it with biovolume measurements by direct microscopy. They concluded that selective inhibition and biovolume measurements were equally effective in measuring the proportion of fungi and bacteria in soils which are either unamended or undergoing rapid changes in metabolism because of substrate amendment, fumigation, or biocidal treatments.

Confirmation of the efficacy of cycloheximide in this soil was sought from the measurement of ergosterol (Table 1). The recovery of ergosterol by the extraction procedure was not significantly different from 100% for all treatments, so the method of Bentham et al. (1992) provided reliable data for this soil. The concentration of ergosterol in the cycloheximide treatment remained constant during the incubation. Up to the 10-h extraction time, the ergosterol concentration in the cycloheximide treatment was not significantly different from the control or streptomycin treatments. However, at the 24-h extraction time, the ergosterol concentration in the cycloheximide treatment was significantly less than the control and streptomycin treatment, which were not different from each other. Cycloheximide prevented the fungal biomass from proliferating up to 24 h confirming the results for respiration (Fig. 1). The reason why all treatments had the same ergosterol concentration up to 10 h was probably because the extraction method did not distinguish between live and dead fungal biomass. It is generally recognized that ergosterol is an indicator of live fungal biomass (Grant and West, 1986; West et al., 1987), but ergosterol degradation following fungal death takes many days. Davis and Lamar (1992) found >95% reduction in ergosterol content within 2 wk following fumigation and death of fungal cells. Nakas and Klein (1979) found that 20 to 70% of all fungal fractions were decomposed in a week. As the incubation time was only up to 24 h in our study, it was very unlikely that ergosterol in dead fungal cells would have been degraded to a significant extent.


View this table:
[in this window]
[in a new window]
  		Table 1. Effect of inhibitors on ergosterol content and recovery during glucose-induced respiration in soil.

 
The Contribution of Bacteria and Fungi to Nitrous Oxide Flux
The flux of N2O into the headspace was unaffected by 15N-labeling, so mean values averaged over all N treatments are presented in Fig. 2 . The flux of N2O in the cycloheximide treatment was insignificant at all times during the incubation, whereas the flux of N2O in the control and streptomycin treatments increased during the incubation. A valid assessment of the relative contribution of bacteria and fungi to N2O flux can be made during the lag phase of growth from 3 to 10 h. On average, the percentage of inhibition of the N2O flux during the lag phase of growth was 89% by cycloheximide and 23% by streptomycin.



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 2. Effect of inhibitors on N2O flux averaged across ammonium nitrate labeling treatments. (LSD value is for comparing any two means at P = 0.05.)

 
The equations of Arah (1997) were used to calculate the fraction of the N2O flux which was derived from the labeled NO3 pool and the 15N mole fraction (aD) of that pool (Table 2). When the NO3 pool was labeled, values of d'D were not significantly different from unity, so all of the N2O was derived from the NO3 pool. Also when the NO3 pool was labeled, the average value for aD, was 0.54. This agrees well with the calculated time-zero value for the enrichment of the labeled NO3 pool (55 atom%, Table 3).


View this table:
[in this window]
[in a new window]
  		Table 2. Effect of inhibitors on the fraction of N2O derived from the labeled nitrate pool , and the 15N mole fraction of that pool (aD), during the lag phase of microbial growth with differentially and doubly labeled ammonium nitrate.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Effect of inhibitors on enrichment of the ammonium, nitrite, and nitrate pools during glucose-induced respiration in the presence of differentially and doubly labeled ammonium nitrate.

 
Recent field studies on the same soil have shown that N2O emissions from fertilizer NO3 were greatly increased in the first few days after application of cattle slurry and that the N2O was being produced by NO3 reduction (Stevens and Laughlin, 2001b). The laboratory study suggests that under such denitrifying conditions in the field, the N2O produced could be largely due to fungi.

The Contribution of Bacteria and Fungi to Dinitrogen Flux
Generation of N2 was indicated by data for {Delta}29R and {Delta}30R (Fig. 3) . For our IRMS system, the between batch limit of detection at the 95% confidence interval is 8.0 x 10-6 for {Delta}29R and 3.2 x 10-7 for {Delta}30R (Stevens and Laughlin, 2001a). Detectable values for {Delta}29R and {Delta}30R were therefore observed when the NO3 pool was labeled, so the labeled N2 was associated with NO3 reduction. In the cycloheximide treatment, all values for {Delta}29R and {Delta}30R were less than the limit of detection. In the control and streptomycin treatments, the production of N2 increased with time with the values for {Delta}29R and {Delta}30R being well above the limit of detection at 24 h. These observations at 24 h were used to calculate the enrichment of the pool from which the N2 was derived (15XN) (Table 4) by the equations of Mulvaney and Boast (1986). The 15XN values were very different from those for aD, the enrichment of the pool from which N2O was derived (Table 2). Neither were the 15XN values similar to the enrichment of any measured mineral N pool (Table 3). Either the assumptions made for the calculation of 15XN are invalid in this study or a process other than denitrification is producing the N2.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 3. Effect of inhibitors and labeling of ammonium nitrate on enrichment of N2 as indicated by {Delta}29R (a, b, c) and {Delta}30R (d, e, f). (LSD value is for comparing any two means at P = 0.05.)

 

View this table:
[in this window]
[in a new window]
  		Table 4. Enrichment of the labeled pool from which N2 was derived (15XN), N2 flux calculated by three methods at 24 h when nitrate was labeled in the control and streptomycin treatments, and N2O fluxes related to N2 fluxes.

 
The main assumption when using the equations of Mulvaney and Boast (1986) is that only one uniformly labeled denitrifying pool exists. Multiple pools of varying enrichment would have resulted in 15XN being overestimated (Arah, 1992; Boast et al., 1988) instead of being underestimated, as occurred in this study. There was little chance of multiple pools existing in this study. The labeled NO3 was applied in solution and mixed with the soil. The amount of added label (6.74 µmol N g-1 oven-dry soil) exceeded the native soil NO3 (0.60 µmol N g-1 oven-dry soil) so that the calculated enrichment assuming total mixing was 55 atom% at time zero. On average the enrichment of the extracted NO3 pool was 55 atom% (Table 3) and the enrichment of the source pool for N2O (aD) was 54 atom % (Table 2). A process other than conventional denitrification or nitrifier denitrification (Poth and Focht, 1985) must have been producing at least some of the labeled N2. Values of 15XN were significantly lower in the 15NH154NO3 treatment than in the NH154NO3 treatment (Table 4). Somehow, 15N from the labeled NH4 pool contributed to the enrichment of the N2. The extent of the contribution was insufficient to produce detectable amounts of labeled N2 in the 15NH4NO3 treatment (Fig. 3a,d).

Chemodenitrification and codenitrification can produce N2 by mechanisms different than denitrification. For conventional denitrification with a source pool of 55 atom%, the ratio of {Delta}29R/{Delta}30R should be 1.63 at all fluxes. In this study, the average ratio of {Delta}29R/{Delta}30R was 17 for data from the control and streptomycin treatments at 24 h. When codenitrification occurs, each N2 molecule is a hybrid of one atom from an enriched pool and one atom from a natural abundance pool, so the ratio of {Delta}29R/{Delta}30R is 272 (Clough et al., 2001). Fungi have been shown to produce N2 by codenitrification (Shoun et al., 1992). Chemodenitrification is also capable of producing hybrid N2 if labeled NO2 from the reduction of labeled NO3 reacts with organic N compounds at natural abundance (Chalk and Smith, 1983). The pH of the soil used in this incubation was 6.3, which is too high to favour chemodenitrification. Since cycloheximide inhibited the production of labeled N2, codenitrification by fungi is considered the most likely process producing hybrid N2 in the control and streptomycin treatments. Values for 15XN were significantly lower in the streptomycin treatment than in the control for the NH154NO3 and 15NH154NO3 treatments (Table 4). Streptomycin inhibits bacterial growth, hence the increased contribution of fungi to N2 production by codenitrification could have increased {Delta}29R relative to {Delta}30R and decreased 15XN. Codenitrification could have occurred along with denitrification in this study.

The N2 flux as calculated by Method 1 is an overestimate of the true N2 flux when 15XN is underestimated. Using Method 2, the flux of N2 due to denitrification was more than 20 times smaller than the flux calculated using Method 1 (Table 4). The most reliable estimate of N2 flux is considered to be that calculated by Method 3 as the sum of denitrification and codenitrification. Codenitrifcation was the dominant process producing N2 accounting for 92% of the total N2 flux at 24 h (Table 4).

The flux of N2O was always greater than the flux of N2, the mole fractions of N2O in the control and streptomycin treatments for measurements at 24 h averaging 0.7 over both labels (Table 4). There was no evidence for N2O production by codenitrification although fungi are capable of this process (Tanimoto et al., 1992). In a field study with the same soil, similar values for the mole fraction of N2O were measured immediately after the addition of glucose (Stevens et al., 1998). Many fungi lack N2O-reductase (Shoun et al., 1992), therefore if a soil is dominated by fungi the mole fraction of N2O would be expected to be close to unity.

Mechanism of Dinitrogen and Nitrous Oxide Production by Fungi
The NO2 pool sizes increased with time in the control and streptomycin treatments but not in the cycloheximide treatment until 24 h (Table 5). The enrichment of the NO2 pool (Table 3) showed that it was predominantly derived from NO3 reduction. Fungi were therefore capable of reducing NO3 to NO2, hence supplying an enriched NO2 pool for codenitrification. The mechanism for NO2 reduction by codenitrification has been proposed as a nitrozation reaction, in which NO3-reductase catalyses the transfer of a nitroso group from NO2 to a nucleophilic N compound such as amino acid (Shoun et al., 1992; Tanimoto et al., 1992). The reaction is pH-dependent, the production of N2 being greater at pH 8 than at pH 6 (Shoun et al., 1992).


View this table:
[in this window]
[in a new window]
  		Table 5. Average effect across ammonium nitrate labeling treatments of inhibitors on the size of the ammonium, nitrite, and nitrate pools during glucose-induced respiration.

 
In this study, the soil was incubated with a headspace of normal atmosphere but low O2 microsites could have been present within soil aggregates. The denitrification of NO3 by fungi requires a minimal initial amount of O2, and is completely repressed by excess O2 (Shoun et al., 1992; Zhou et al., 2001). Denitrification by the fungal species F. oxysporum required less O2 with NO2 than with NO3 (Usada et al., 1995; Zhou et al., 2001). Dinitrogen production from NO2 by codenitrification has been reported to occur in completely anaerobic conditions (Shoun et al., 1992). There is mounting evidence that denitrification by fungi and O2 respiration share the mitochondrial respiratory chain and simultaneously occur under specific conditions of O2 stress. Denitrification by fungi has been shown to be inhibited by respiration inhibitors (Shoun and Tanimoto, 1991) and mitochondria prepared from denitrifying cells of F. oxysporum contained both dissimilatory NO3 reductase and cytochrome c oxidase activities (Kobayashi et al., 1996).

Fungal Dominance of Mineral Nitrogen Transformations
The mean values for the size (Table 5) and enrichment of the NH4 pool (Table 3) in the 15NH4NO3 treatment were used to calculate the rates of mineralization and consumption of NH4 (Table 6). Likewise the mean values for the size (Table 5) and enrichment (Table 3) of the NO3 pool in the NH154NO3 treatment were used to calculate the rates of nitrification and consumption of NO3. Confirmation that fungi dominated mineralization-immobilization turnover was provided by data from the pool sizes and enrichment of the 15NH4NO3 treatment. Immobilization was four times greater than mineralization because of the added glucose. Streptomycin had no effect on the immobilization of NH4 whereas cycloheximide decreased the immobilization of NH4 by 88%, indicating that fungi dominated immobilization.


View this table:
[in this window]
[in a new window]
  		Table 6. Effect of inhibitors on gross N transformation rates calculated during 24 h of glucose-induced respiration.

 
Nitrification rates were low in comparison with immobilization rates. Streptomycin and cycloheximide decreased nitrification by 22 and 78%, respectively (Table 6), indicating that fungi also dominated nitrification. The importance of the fungal contribution to nitrification has been reported for woodland soils (Castaldi and Smith, 1998; Killham, 1987, 1990; Schimel et al., 1984). Fungi are the most numerous and efficient heterotrophic organisms able to produce NO3 in soil (Odu and Adeoye, 1970). Heterotrophic nitrification can be from NH4 or from organic N (Paul and Clark, 1996). In this study, the inorganic pathway can be inferred from the enrichment of the NO3 in the 15NH4NO3 treatment (Table 3). The NO3 pool remained unlabeled in the cycloheximide treatment but became significantly enriched with time in the control and streptomycin inhibitor treatments. Castaldi and Smith (1998) suggested the possible role of ammonia monooxygenase in fungal metabolism. The rates of NO3 consumption were small compared with NH4 consumption (Table 6) and could be due to assimilatory or dissimilatory pathways. Fungi can assimilate NO3 but it is not utilized unless the cells lack a favoured N source, such as NH4, glutamine, or glutamate (Marzluf, 1997).


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 Substrate Induced Respiration...
 Headspace Sampling and Soil...
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
If any process other than the sequential reduction of NO3 is producing N2O or N2, there are implications for the measurement of denitrification by the 15N gas-flux method. In this study, N2O was produced by NO3 reduction but N2 was produced predominantly by codenitrification. If codenitrification occurs, the conventional equations using {Delta}29R and {Delta}30R will underestimate the enrichment of source pool for N2 (15XN) and overestimate the flux of N2. Using only {Delta}30R and assuming that 15XN is the same as the enrichment of the denitrifying source pool calculated from N2O data (aD) will underestimate the N2 flux. If the calculated 15XN value is much less than the value calculated for aD, codenitrification may be occurring and producing some N2. The flux of N2 would then have to be apportioned between codenitrification and denitrification to obtain a reliable estimate of N2 flux.

In 15N mass-balance studies a portion of the applied 15N is often not recovered (Myrold, 1990). This portion is assumed to be the amount of N lost by denitrification. If codenitrification was occurring instead of denitrification each molecule of N2 would be formed from a 15N atom in the labeled pool and a 14N atom in the natural abundance pool. The amount of 15N not recovered would only be half of the total N2 lost from the soil.

The occurrence of fungal denitrification is of ecological significance as N2O is the dominant gaseous product. As fungi have the ability to perform denitrification and O2 respiration simultaneously in a range of O2-stress conditions, the potential exists for fungi to produce N2O in a wider range of soil aeration conditions than bacteria, which need anaerobic conditions to denitrify. Fungi are widely distributed in soils and water (Zvyagil'skaya et al., 1996), hence the potential exists for fungi to make a significant contribution to the global N2O budget.


   ACKNOWLEDGMENTS
 
We thank Tim Clough (Lincoln University, New Zealand) for his assistance with deriving equations for the calculation of codenitrification and Michael Nicholson (Department of Agriculture and Rural Development, UK) for his technical assistance.

Received for publication August 2, 2001.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 Substrate Induced Respiration...
 Headspace Sampling and Soil...
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
S. Haile-Mariam, H. P. Collins, and S. S. Higgins
Greenhouse Gas Fluxes from an Irrigated Sweet Corn (Zea mays L.)-Potato (Solanum tuberosum L.) Rotation
J. Environ. Qual., May 1, 2008; 37(3): 759 - 771.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
V. D. Zheljazkov, T. Astatkie, C. D. Caldwell, J. MacLeod, and M. Grimmett
Compost, Manure, and Gypsum Application to Timothy/Red Clover Forage
J. Environ. Qual., October 27, 2006; 35(6): 2410 - 2418.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
S. Menendez, P. Merino, M. Pinto, C. Gonzalez-Murua, and J. M. Estavillo
3,4-Dimethylpyrazol Phosphate Effect on Nitrous Oxide, Nitric Oxide, Ammonia, and Carbon Dioxide Emissions from Grasslands
J. Environ. Qual., May 31, 2006; 35(4): 973 - 981.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
J. E.T. McLain and D. A. Martens
Moisture Controls on Trace Gas Fluxes in Semiarid Riparian Soils
Soil Sci. Soc. Am. J., February 2, 2006; 70(2): 367 - 377.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (41)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Laughlin, R. J.
Right arrow Articles by Stevens, R. J.
Right arrow