N₂O production, a widespread trait in fungi

Abstract

N₂O is a powerful greenhouse gas contributing both to global warming and ozone depletion. While fungi have been identified as a putative source of N₂O, little is known about their production of this greenhouse gas. Here we investigated the N₂O-producing ability of a collection of 207 fungal isolates. Seventy strains producing N₂O in pure culture were identified. They were mostly species from the order Hypocreales order—particularly Fusarium oxysporum and Trichoderma spp.—and to a lesser extent species from the orders Eurotiales, Sordariales, and Chaetosphaeriales. The N₂O ¹⁵N site preference (SP) values of the fungal strains ranged from 15.8‰ to 36.7‰, and we observed a significant taxa effect, with Penicillium strains displaying lower SP values than the other fungal genera. Inoculation of 15 N₂O-producing strains into pre-sterilized arable, forest and grassland soils confirmed the ability of the strains to produce N₂O in soil with a significant strain-by-soil effect. The copper-containing nitrite reductase gene (nirK) was amplified from 45 N₂O-producing strains, and its genetic variability showed a strong congruence with the ITS phylogeny, indicating vertical inheritance of this trait. Taken together, this comprehensive set of findings should enhance our knowledge of fungi as a source of N₂O in the environment.

Terrestrial ecosystems are a major source of nitrous oxide (N₂O)¹,², a so-called greenhouse gas also commonly known as laughing gas. Although it has received much less attention than CO₂, the 100 year global warming potential of N₂O is 298 times greater than that of CO₂ due to the much longer half life of N₂O³. There is also growing concern over nitrous oxide concentrations because, following the reduction of chlorine- and bromine-containing halocarbons by the Montreal Protocol, N₂O has become the main ozone-depleting substance emitted to the stratosphere⁴.

Nitrous oxide emissions are mostly due to two microbial processes: nitrification and denitrification. Nitrous oxide is a by-product of the first step of nitrification, the oxidation of ammonia to nitrite⁵. In contrast, N₂O is either an intermediate or the end product of the denitrification cascade, which consists in the reduction of nitrate or nitrite into nitric oxide, nitrous oxide and dinitrogen⁶. Sixty-two percent of the total global N₂O emissions are from natural and agricultural soils (6 and 4.2 Tg N yr⁻¹, respectively⁷), and denitrification is traditionally considered as the main source of these emissions⁸.

It is well known that denitrification is widespread among prokaryotes—indeed, the ability to denitrify has been observed in more than 60 bacterial and archaeal genera⁹. Moreover, eukaryotes such as fungi in soils¹⁰ or foraminifers in aquatic environments¹¹,¹² are also capable of denitrification. Characterization of the fungal denitrification ability in Fusarium oxysporum and Cylindrocarpon tonkinense has shown that this reductive process was performed via a copper-containing nitrite reductase (NirK) and cytochrome P450 nitric oxide reductase¹⁰. However, no nitrous oxide reductase has been identified in fungi and N₂O is the end product of denitrification in the few characterized fungal strains¹³,¹⁴. By using fungal or bacterial inhibitors to distinguish the microbial origin of N₂O, previous studies have reported that fungi could contribute up to 18% of potential denitrification¹⁵ and be significant N₂O producers in some terrestrial systems¹⁶,¹⁷. Despite the importance of fungi in several soil functions, such as organic matter decomposition¹⁸ and primary production through symbiotic or pathogenic relationships with plants¹⁹, the production of N₂O by fungi has only been studied in a limited number of strains¹⁴,²⁰. To what extent this trait is conserved amongst fungi remains unknown, but understanding the microbial sources of this greenhouse gas will be crucial for selecting mitigation strategies. Here, we screened a collection of 207 fungal strains belonging to 9 classes and 23 orders to determine the prevalence of the N₂O-producing capacity among fungi. We further characterized the initial and end-products of denitrification of the N₂O production-positive strains in pure culture and determined their N₂O isotopic signature. Positive fungal strains were also inoculated into pre-sterilized arable, forest and grassland soils in order to verify their ability to produce this greenhouse gas in soil. Finally, we studied the phylogeny of the nirK gene, which encodes the copper-containing nitrite reductase using newly developed primers, and investigated the relationships between the nuclear ribosomal internal transcribed spacer (ITS) region and nirK phylogeny, N₂O production rates and N₂O isotopic signatures.

Results and Discussion

To assess how the N₂O producing ability is distributed within fungi, 207 fungal strains comprising 23 orders and 54 genera were screened by incubating the strains in liquid culture under conditions that were previously reported to favour fungal denitrification²¹. The strains were selected to cover the largest possible fungal diversity within the Microorganisms of Interest for Agriculture and Environment collection (MIAE) (INRA, Dijon, France), which is dedicated to soil microbial diversity. At the end of the incubation, differences in the pH of the media were observed between strains. Since N₂O can also be produced by chemical denitrification at low pH²², abiotic N₂O production from nitrite was evaluated in sterile media with a pH gradient, and strains were scored positive when the N₂O concentrations in the headspace were higher than those in the sterile flasks at the same pH (Fig. S1). When nitrite was used as an electron acceptor, more than a third of the strains were capable of producing N₂O, with activities ranging from 0.5 ± 0.1 to 60.0 ± 36.0 mg N₂O-N g⁻¹ dry fungal biomass (Fig. 1). The N₂O-producing activities were much lower when nitrate was used as an electron acceptor (<0.1 mg N₂O-N g⁻¹ fungal biomass; F = 108.55, P < 0.0001), supporting previous studies showing that nitrite rather than nitrate is preferable for fungal denitrification²³. No difference was observed when incubating the positive strains with and without acetylene, indicating that the fungi does not reduce N₂O, which was also in accordance with previous studies²⁴. Accordingly, amplification of the nosZ gene using various primer sets²⁵,²⁶ was not successful (data not shown). The high proportion of N₂O-producing fungal strains observed in our study contrasts with previous studies in which only 1% to 10% of examined bacteria were capable of denitrification based either on culture-based, direct-molecular approaches or genome analyses²⁷,²⁸,²⁹. However, the maximum percentage of nitrogen recovered as N₂O from nitrite in our work was about 38%, and most of the fungal strains reduced between 3% and 10% of the nitrite, while denitrifying bacteria are capable of reducing at least 80% of soluble nitrogen into gas³⁰. Nonetheless, lower percentages were also reported for denitrifying bacteria such as Bacillus species, with ranges between 3.5% and 13.2% of the nitrate reduced to gaseous nitrogen after 48 h of growth³¹. Although we cannot rule out that growing the strains in different media or conditions may have resulted in different rates of nitrogen reduction, our incubation experiment using the standard media and conditions showed that N₂O production is common within the fungal kingdom.

Figure 1. N₂O production by the positive fungi strains with (white) and without (black) 10% C₂H₂ in the headspace.

Error bars indicate the standard deviation (n = 3). Both strain names and MIAE numbers are indicated.

Fifty out of the 70 positive strains belonged to the Hypocreales order, and Fusarium and Trichoderma were the main Hypocreales genera observed. Interestingly, many of the Fusarium strains identified as N₂O-producers were Fusarium oxysporum. Even within this species, the production rate was highly variable, ranging from 2.8 to 34.7 mg N₂O-N g⁻¹ dry fungal biomass. This species was reported to be one of the dominant fungal taxa in several studies³², and accounted for up to 43% of the ITS pyrosequencing dataset retrieved from Mediterranean soils³³. F. oxysporum includes non-pathogenic and pathogenic strains, with the latter causing disease to a broad range of host plants, but no plant-based bioassay has been conducted on the tested strains to discriminate pathogenic and non-pathogenic F. oxysporum. In any case, the high number of N₂O-producing F. oxysporum individuals suggests this species is involved in greenhouse gas emissions, and therefore are potentially detrimental in terms of both primary production and climate regulation. In both Trichderma and Fusarium species capable of living in plant tissue, respiration of nitrogen oxides is likely due to the adaptation to hypoxic conditions. Indeed, the oxygen concentration near or within plant tissue is low (<1%)³⁴, and several studies have reported that it is critical to the fungal infection of plants that the infecting fungus possess strategies to overcome hypoxia³⁵. The other positive strains identified in this work belong to the Eurotiales (8), Sordariales (5), Chaetosphaeriales (3), Mucorales (1), Pleosporales (1), Glomerellales (1) and Ophiostomatales orders (1). Sordariales has also been reported as a dominant order in agricultural soil, while the order Pleosporales is frequently detected in that environment³². Collectively, our results indicate that N₂O production is a common trait in fungal taxa that are frequently abundant in soils, although this feature seems to be more a strain-specific than a species-specific trait. It also underlines the importance of quantifying the fungal contribution to terrestrial N₂O emissions.

Analysis of the N₂O isotopomer ratios (relative abundance of ¹⁴N¹⁵N¹⁶O or ¹⁵N¹⁴N¹⁶O to that of ¹⁴N¹⁴N¹⁶O) has been proposed as a powerful method to obtain more information on the sources of this greenhouse gas³⁶. However, the site preference (SP) values due to fungal N₂O production have only been determined in 2 strains (F. oxysporum and C. tonkinense)³⁷, and a more comprehensive analysis of the variability of the fungal ¹⁵N site preference is required for a robust distinction between the bacterial and fungal contributions to N₂O emissions. The SP values of the 67 fungal strains tested herein varied from 15.8–36.7‰ with an average value of 30.0 ± 4.8‰. Our results are partly consistent with the previous study reporting a positive SP of 36.9–37.1‰³⁷. However, we found a larger range of variation, with SP values as low as 15.8 ± 2.6‰ for Penicillium melinii and as high as 36.7 ± 2.2‰ for F. sambucinum. These strains with low SP values often had more acidic conditions at the end of the incubation (pH 3–5; Fig. S1). Therefore, we cannot rule out that these lower SP values resulted from the higher contribution of abiotic N₂O production under more acidic conditions³⁸. Interestingly, a significant taxa effect (P < 0.01 by Student's t-test) with lower SP values was observed for the Penicillium strains compared to the other fungal genera. We thus confirmed that the N₂O isotopic signature can be used to determine which organisms, i.e., fungi or bacteria, are producing N₂O by denitrification because of their distinct SP (about 0–10‰ for bacterial denitrification³⁹). However, it will be difficult to distinguish N₂O emissions from nitrification and fungal denitrification in the environment, since SP values for the N₂O produced by both bacteria and archaea during ammonia oxidization ranged between 13.1 and 30.2‰⁴⁰,⁴¹. Since fungi are microaerophilic denitrifiers, our results suggest that a stable isotopic approach alone is not enough to decipher whether nitrification or fungal denitrification is contributing to N₂O emissions in environments where fungi are abundant.

To further confirm that fungi could actually produce N₂O in soil and not only in liquid culture, 15 of the positive strains were selected based on their high activity in pure culture and/or their taxonomic affiliation for inoculation into an arable, a grassland and a forest pre-sterilized soils. After allowing soil colonization by the inoculated fungal strains for a month, nitrite was added to induce denitrification and the concentration of N₂O in the headspace was measured after 2, 4 and 7 days. Seven strains belonging to the Trichoderma, Fusarium, Penicillium and Phialocephala genera produced N₂O in at least one of the soils, up to a maximum amount of 82.1 ng N₂O-N g⁻¹ h⁻¹ (Fig. 2), which was significantly higher (F = 15590, P < 0.0001) than the amount produced by chemical denitrification in the non-inoculated sterile soils. This is comparable to previous studies focusing on fungal contribution¹⁵,¹⁶,¹⁷,⁴² or net N₂O production from soil⁴³,⁴⁴,⁴⁵. In our study, the highest amount of N₂O produced was observed in the forest soil for Fusarium verticillioides and F. dimerium, while the grassland and the arable soils were the soils with the highest emissions for Trichoderma harzianum and Phialocephala spp. on one hand, and F. oxysporum f. sp. lini, Metarhizium anisopliae on the other. These significant strain-by-soil effects (F = 4.69, P < 0.0001) likely reflect differences in fungal nutrient requirements and/or preferences for different soil physico-chemical characteristics. Indeed, previous studies demonstrated that several soil physico-chemical parameters, such as extractable P concentrations, C/N ratio⁴⁶, pH, sand content or litter cover⁴⁷, can affect the soil fungal communities or fungal biomass⁴⁸. The fungal N₂O-producing capacity observed in our study could also be influenced by biotic interactions occurring in the natural ecosystems between fungi or with bacteria and other organisms.

Figure 2. N₂O production by 15 selected fungi strains inoculated in 3 different sterile soils.

NO₂⁻ was used as the electron acceptor. The strains were incubated for 7 days, and headspace N₂O concentration were measured 3 times (2, 4 and 7 days after NO₂⁻ addition). Code for the soils: Black circle: Sweden (Slogaryd), forest soil; grey triangle: France, arable soil; black rectangle: The Netherlands, grassland soil. Code for the fungi: a: Trichoderma harzianum (MIAE00042); b: Fusarium verticillioides (MIAE00306); c: Penicillium adametzii (MIAE01008); d: F. oxysporum f. sp. lini (MIAE00347); e: F. dimerum (MIAE00598); f: Metarhizium anisopliae (MIAE00953); g: Chaetomium sp. (MIAE00985); h: T. harzianum (anamorph) (MIAE01011); i: Leptosphaeria sp. (MIAE01060); j: T. tomentosum (MIAE00031); k: Fusarium sp. (MIAE01519); l: Clonostachys candelabrum (MIAE00941); m: Phialocephala sp. (MIAE00968); n: Colletotrichum coccodes (MIAE01515); o: Aspergillus sp. (MIAE01518);); p: No fungi control. The error bar represents the standard deviation (n = 3).

Both copper-containing nitrite reductase (encoded by the nirK gene) and P450nor (nitric oxide reductase) are key enzymes involved in fungal denitrification. However, P450nor belongs to a superfamily of proteins that are widely distributed among fungi and known to be involved in a wide variety of physiological reactions¹⁰, which prevents the use of the corresponding genes as molecular markers to target denitrifying fungi. To date, several primers targeting the nirK denitrification gene have been described in the literature, but none of them was designed to amplify denitrifying eukaryotes²⁵,⁴⁹. Despite the low number of fungal nirK sequences available in the databases (less than 30) and the high diversity of the tested fungal strains, the amplification of the fungal nirK denitrification gene was successful in 45 out of 70 strains using our newly designed primer set EunirK-F1 and EunirK-R2. This supports our findings that these fungal strains are capable of denitrification and that N₂O was not produced by other processes. Notably, when used to amplify DNA extracted from soil, our primer set also amplified the bacterial nirK in part due to the lower proportion of fungi in soil compared to bacteria (data not shown). A phylogenetic tree was constructed using these fungi nirK sequences and bacterial nirK sequences from available databases (Fig. 3). Our fungal nirK sequences clustered with the other fungal nirK homologues retrieved from the database, and were distinct from the bacterial nirK sequences. The phylogeny also shows that fungal nirK sequences are closer to other eukaryotic sequences (amoeba, protozoa or green alga) than to bacterial ones, indicating that fungal nirK sequences branched from bacterial nirK sequences at a very early stage of their evolutional history, as suggested by Kim et al.⁵⁰. In addition, we also found a strong congruence between the ITS and the nirK phylogenies (Fig. S2), indicating a vertical inheritance of nirK genes. Interestingly, we found no correlation between the genetic distance of the nirK genes and the N₂O production rates. Similarly, weak or no relationships were observed between the bacterial genotypes and denitrification phenotypes in previous studies³¹,⁵¹. Phenotypic convergence within similar ecological niches of distantly related organisms can lead to such a discrepancy between genetic and phenotypic distances. In depth investigation of the ecology of denitrifying fungi would undoubtedly help clarify which environmental factors lead to a convergent denitrification phenotype.

Figure 3. Neighbour-joining phylogenetic tree of nirK amino acid sequences constructed by Clustal W with 1000 bootstrap samplings.

Strain names in bold indicate the sequences obtained in this study. The numbers in parentheses indicate the number of the strains. Bootstrap values greater than 75% are indicated as black circles.

In conclusion, the analysis of a vast collection of fungi showed that N₂O production is a common and widespread trait in fungi. Nitrite instead of nitrate was the preferred substrate, while N₂O was always the end-product of denitrification. We showed that the range of variation of the N₂O isotopic signature was taxa-dependent and larger than previously reported, with values as low as 15%. Inoculation of 15 strains into previously sterilized arable, forest and grassland soils demonstrated the ability of fungi to contribute to soil N₂O emissions with fluxes potentially as high as those reported in natural soils. Further studies are clearly warranted to elucidate the significance of denitrification in fungi and its consequences for N₂O emissions.

Methods

Fungal strains and in vitro incubation experiments

The fungal strains tested in this study were previously isolated from agricultural soils or plant roots. The details of the isolating procedures have been described previously⁵². The strains were purified by single-spore isolation and preserved in the MIAE collection (INRA, Dijon, France). Fungal strains were first cultured under an aerobic condition in a 147 ml plasma flask containing 50 ml of liquid malt medium (pH 7.5). The flasks were incubated with a rotator shaker at 25°C, 120 rpm. After 3 days to 2 weeks depending on their growth, 50 ml of liquid malt medium with 10 mM NaNO₂ (pH 7.5) was added to the plasma flasks, which were capped with a butyl rubber stopper. The headspace gas was then replaced by pure N₂ gas and 1 ml of pure O₂ to obtain microaerobic conditions. The flasks were incubated again at 25°C, 120 rpm for one additional week. The incubations were performed in triplicate flasks for each fungal strain. After the second incubation round, 0.5 ml of headspace gas was sampled and N₂O concentrations were determined by gas chromatography. Three previously characterized strains²⁴,⁵³,⁵⁴, F. oxysporum MT811 (JCM11502), Cylindrocarpon lichenicola (NBRC30561) and Aspergillus oryzae RIB40 (NBRC100959), were used as positive controls. At the end of the incubation, mycelia were collected and dried for 24 h at 105°C for fungal biomass determination, and the pH of the medium was determined by using a commercial electrode.

Strains capable of producing N₂O were also incubated as described above, but with NO₃⁻ as electron acceptor (NaNO₃ at a final concentration of 5 mM), and biomass and pH determination were performed as described above. Gas measurements were done with and without 10% C₂H₂ gas in the headspace⁵⁵ to verify whether or not they were capable of reducing N₂O into N₂.

Fungal N₂O production in sterile soil

The physical and chemical parameters of the three soils used in this study are described in Table 1. Triplicate samples were collected from the top 10 cm of three different soils, sieved to <2 mm and sterilized by γ -radiation (35 kGy; Conservatome, Dagneux, France).

Table 1. Soil characteristics and properties across a range of ecosystem types.

Country	Description	Clay	Loam	Sand	Water content	WHC		Total-C	Total-N		OM
		%	%	%	g g−1 soil DW	g g−1 soil DW	pH	g kg−1 soil DW	g kg−1 soil DW	C:N ratio	g kg−1 soil DW
France	Non-irrigated arable land	18.4	66.4	15.1	0.11	0.50	6.6	10.2	0.97	10.6	17.7
The Netherlands	Permanent Grassland	3.4	2.7	93.8	0.17	0.44	5.1	20.3	1.25	16.3	35.1
Sweden	Coniferous forest	21.4	15.9	62.7	1.51	2.54	4.3	253.0	14.2	17.8	438.0

WHC: water holding capacity; OM: organic matter; DW: dry weight.

Denitrifying strains were incubated with liquid malt medium (400 mL) in 1 L flasks at 25°C, 120 rpm for 7 to 14 days so that enough biomass was obtained. Liquid cultures were centrifuged (12,000 rpm for 10 min) in 50 mL tubes and washed twice with 30 mL of sterile physiological water (0.9% NaCl). For each soil, fungal pellets were resuspended with sterile water and inoculated into two series of triplicate microcosms containing the γ-ray sterilized soil (5 g in 147 mL sterile bottles). Soil moisture was adjusted in order to obtain a water holding capacity (WHC) of 60% after the inoculation, and all the inoculated soil microcosms were incubated for 3–4 weeks at room temperature to allow soil colonization by the fungi. After this pre-incubation, 2 mL of NaNO₂ solution (10 mM) was added and the soil moisture was adjusted at a WHC of 90%. In half of the replicated microcosms (three for each soil), 10 mL of the ambient air in the headspace was replaced with 10 mL C₂H₂, and all bottles were incubated at 25°C for one additional week. Production of N₂O was also monitored in all three sterile soils without inoculation of fungi (negative controls).

Nitrous oxide emission measurements

Nitrous oxide production by fungal pure culture was determined by analysing the gas samples collected in the headspace gas using a gas chromatograph (TRACE GC Ultra; Thermo) equipped with an electron capture detector (GC-ECD).

In the inoculated soil microcosms, gas samples were collected 3 times (2, 4, and 7 days after NO₂ addition) from the headspace and were analysed with GC-ECD to determine the N₂O concentrations. The N₂O concentrations were analysed by ANOVA using the GLM procedure in SAS⁵⁶. For the liquid medium culture experiment, strain and substrate (NO₂ or NO₃) were used for the fixed effect. For the sterilized soil incubation experiment, strain and soil were used for the fixed effect and strain*soil was included in the model. Tukey's multiple comparisons test was used to separate the means. The significance level was 0.05.

Nitrous oxide isotopic signature measurements

Ten ml of each headspace gas sample was taken at the end of the incubation experiments and stored in a pre-evacuated vial. The samples were then introduced into a gas chromatograph-isotope ratio mass spectrometer (GC-IRMS) (MAT 252; Thermo Fisher Scientific K.K., Yokohama, Japan) system as described elsewhere to measure the N₂O isotopomer ratios⁵⁷. Site-specific N isotope analysis in N₂O was conducted using ion detectors that had been modified for the mass analysis of fragment ions of N₂O (NO⁺) containing N atoms in the centre positions of N₂O molecules, whereas the bulk (average) N and oxygen isotope ratios were determined from molecular ions⁵⁸. Pure N₂O (purity > 99.999%; Showa Denko K.K., Japan) was calibrated with international standards and used as a working standard for the isotopomer ratios. The notation of the isotopomer ratios is shown below. The measurement precision was typically better than 0.1‰ for δ¹⁵N^bulk and δ¹⁸O, and better than 0.5‰ for δ¹⁵N^α and δ¹⁵N^β.

Here, ¹⁵R^α and ¹⁵R^β respectively represent the ¹⁵N/¹⁴N ratios at the centre and end sites of the nitrogen atoms; ¹⁵R^bulk and ¹⁸R respectively indicate the average isotope ratios for ¹⁵N/¹⁴N and ¹⁸O/¹⁶O. The subscripts “sample” and “std” respectively indicate the isotope ratios for the sample and the standard, atmospheric N₂ for N and Vienna Standard Mean Ocean Water (V-SMOW) for O. We also define the ¹⁵N site preference (hereinafter SP) as an illustrative parameter of the intramolecular distribution of ¹⁵N:

The N₂O concentration was measured simultaneously with the isotopomer ratios by comparing the peak area of the major ion (mass 44 and 30 in molecular ion analysis and fragment ion analysis, respectively) obtained with the sample gas and with a reference gas (349 nL/L N₂O in Air; Japan Fine Products Co., Ltd.)⁵⁷.

Primer design

Full-length nirK nucleotide sequences of fungal genomes were obtained from the Functional Gene Pipeline public database (http://fungene.cme.msu.edu//index.spr). These sequences were aligned using MEGA in order to design fungal specific −nirK primers. The primer sequences were as follows: EunirK-F1 (5′-GGB AAY CCI CAY AAY ATC GA-3′) and EunirK-R2 (5′-GGI CCI GCR TTS CCR AAG AA-3′).

DNA extraction, and PCR amplification of the nirK gene and ITS from positive strains

DNA extraction from the denitrifying fungal cultures was performed using a commercially available DNA extraction kit, DNeasy® Plant Maxi (QIAGEN). The extraction was performed according to the manufacturer's instructions. The purified DNA samples were stored at −20°C until further analysis.

The PCR protocol for the nirK gene was as follows: 10 min at 94°C and 40 cycles consisting of 1 min at 94°C, 30 s at 53°C, and 1 m at 72°C. For amplification of the fungi Internal Transcribed Spacer (ITS) region, the primer sets ITS5 (5′-TCC TCC GCT TAT TGA TAT GC-3′) and ITS4 (5′-GGA AGT AAA AGT CGT AAC AAG G-3′) were used⁵⁹. The PCR protocol was as follows: 10 min at 95°C and 35 cycles consisting of 95°C for 15 s, 30 s at 52°C, and 72°C for 1.5 s. These amplicons were purified and sequenced with the BigDye Terminator version 3.1 cycle sequencing kit (Applied Biosystems) and the ABI Prism 3100 genetic analyser. The nucleotide sequence of the ITS regions of other fungi and bacterial nirK (both class 1 and 2) amino acid sequences were also obtained from the database described above, and the phylogenetic tree was constructed based on the maximum likelihood method using CLUSTALW. Congruence between the ITS and nirK phylogenies was graphically illustrated using the cophyloplot function from the “ape” R package⁶⁰.

Author Contributions

L.P. and C.S. designed the experiments. K.M., A.S. and L.P. wrote the paper. K.M., C.H., M.C.B. and F.B. screened the fungal isolates. V.E.H. verified the fungal isolates. K.M., S.T. and N.Y. worked on the isotopic experiments.

Additional Information

Accession codes: All sequence data were submitted to DDBJ under accession numbers AB904794-AB904830.

Supplementary Material

Supplementary Information

Supplementary Information