Isotope analyses of amino acids in fungi and fungal feeding Diptera larvae allow differentiating ectomycorrhizal and saprotrophic fungi-based food chains

1. Both ectomycorrhizal (ECM) and saprotrophic fungi are fundamental to carbon and nutrient dynamics in forest ecosystems; however, the relative importance of these different fungal functional groups for higher trophic levels of the soil food web is virtually unknown. 2. To explore differences between fungal functional groups and their importance for higher trophic levels, we analysed isotopic composition of nitrogen and carbon in amino acids (AAs) and bulk tissue of leaf litter, fungi and fungal-feeding Diptera larvae. 3. By accounting for isotopic variability of utilized substrates, compound-specific isotope analyses of nitrogen in AAs yielded more realistic results for the trophic

fungal contributions to energy and nutrient fluxes in soil food webs, it is essential to know the trophic position of fungi and their consumers.
Other explanations include fractionation processes during chitin synthesis (Gleixner, Danier, Werner, & Schmidt, 1993), or the utilization of 13 C enriched components of leaf litter, such as cellulose (Hobbie et al., 1999). Similar to saprotrophic fungi, soil invertebrates, including primary and secondary decomposers as well as predators, are enriched in 13 C. This enrichment has been termed 'detrital shift'  and at least in part is due to the preferential utilization of plant compounds enriched in 13 C such as starch or cellulose (Pollierer, Langel, Scheu, & Maraun, 2009).
Compound-specific isotope analysis (CSIA) of amino acids (AAs) is increasingly used to gain more profound insights into trophic positions and basal resources of consumers. Using the difference in δ 15 N values (i.e. 15 N/ 14 N ratios) between so-called 'trophic' and 'source' AAs, trophic positions of consumers can be calculated, with the advantage that the isotopic baseline is recorded in the source AA isotopic signature. Most studies use a canonical single pair of the AAs, that is, glutamate (Glu) and phenylalanine (Phe), as trophic and source AAs, respectively, as their trophic enrichment typically is relatively constant among consumers of different groups and trophic levels (McMahon & McCarthy, 2016). Steffan et al. (2015) suggested that microbes can be treated as trophic analogues of animals, that is, that they occupy a similar trophic position when feeding on the same resource. Therefore, saprotrophic fungi should be positioned one trophic level above the leaf litter they consume. By contrast, estimation of trophic position in mycorrhizal fungi is not as straightforward; they may transfer isotopically depleted ammonium or amino acids to their host plants (Hobbie & Högberg, 2012) and are considerably enriched in bulk 15 N compared to the latter.
Compound-specific isotope analysis of C in essential AAs has been used to differentiate between consumption of bacteria, fungi and plants in an approach termed 'stable isotope fingerprinting' (Larsen, Taylor, Leigh, & O'Brien, 2009;Larsen et al., 2013). The approach is based on the fact that these phylogenetic groups differ in their major biosynthetic pathways for synthesizing AAs, resulting in lineage-specific isotope patterns that can be traced in essential AAs of consumers to distinguish basal resources.
However, to date the microbial datasets used as template for fingerprinting mainly are based on laboratory cultures using artificial AA deficient growth media. It has not been tested yet whether fungi in natural environments fully rely on their own biosynthetic pathways to synthesize essential AAs or whether they take up plant-derived AAs, diluting their phylogenetic 13 C signal in AAs.
Saprotrophic and ECM fungi may differ in AA δ 13 C values since the latter acquire C directly from host plants, mainly in the form of simple sugars, whereas saprotrophic fungi use the more recalcitrant litter as a C source, as is also mirrored in the large differences in bulk C isotopic signatures.
Considerable fluxes of root-derived C into the soil animal food web have been attributed to the consumption of ECM fungi (Pollierer, Dyckmans, Scheu, & Haubert, 2012). However, it remains a riddle whether mycorrhizal fungi are consumed by soil animals in significant amounts. Comparisons of isotopic signatures of soil animals to those of other consumers of saprotrophic micro-organisms, such as mycotrophic plants (Potapov & Tiunov, 2015) or fungus growing termites , suggest that fungal feeders primarily rely on saprotrophic fungi; however, this suggestion needs further verification as there also is evidence that few soil animal taxa are specialized ECM fungal feeders, for example, Protura (Bluhm et al., 2019). Potentially, 13 C fingerprinting may separate consumers of saprotrophic and mycorrhizal fungi, providing the opportunity to distinguish these feeding types.
Diptera larvae of the families Mycetophilidae, Anthomyidae and Limoniidae often feed on fungal sporocarps (Hackman & Meinander, 1979). Mycetophilidae are the largest group of fungal feeding Diptera (Jakovlev, 2012); their larvae feed inside sporocarps of macrofungi. Since Diptera larvae are confined to their fungal host, they represent ideal candidates to investigate trophic fractionation in fungal feeders.
To investigate the trophic position of ECM and saprotrophic fungi, and explore if they and their consumers can be distinguished by using stable isotope fingerprinting, we collected fungal sporocarps and Diptera larvae living and feeding within the sporocarps from deciduous and conifer forests near Moscow, Russia. We analysed bulk stable isotope composition and performed CSIA of AAs in leaf litter, fungi and Diptera, calculated trophic positions and applied linear discriminant function analysis to distinguish between fungal groups based on δ 13 C values of essential AAs. We hypothesized that (a) the trophic position of fungi can be calculated more accurately with CSIA of N compared to bulk δ 15 N analysis, since the signature of the utilized resource is recorded in the source AA, (b) Diptera from within fungal sporocarps are one trophic level above respective fungal sporocarps, (c) ECM and saprotrophic fungi differ in 13 C fingerprints due to differences in C acquisition and (d) Diptera from within sporocarps do not differ in essential AA δ 13 C values from respective sporocarps.

| Study site
Fungal sporocarps (cups/gulls) and Diptera larvae were collected along with fresh leaf litter on September 2016 at Malinki Biological Station, south of Moscow, Russia (55.4595°N, 37.1794°E). Samples were collected within few hectares of an over-ripe coniferous 150to 160-year-old forest dominated by Norway spruce and Scots pine.
Replicate samples of fungi and Diptera larvae were taken from individual sporocarps (Table 1). Leaf litter was collected close to the respective fungi.
Diptera larvae were collected from within the fungal sporocarps.

| Bulk stable isotope and C-to-N ratio analyses
For C and N bulk isotope and C-to-N ratio analyses, fungi and leaf litter were dried at 60°C for 24 hr and ground to powder. Appropriate amounts (c. 2 mg for leaf litter and 0.5-1.0 mg for fungi) were weighed into tin capsules and stored in a dessicator until analysis. Diptera larvae (between 0.5 and 0.9 mg dry weight) were also weighed into tin capsules and dried at 60°C for 24 hr. Stable isotope and C-to-N ratios of fungi, Diptera, and leaf litter were determined using a coupled system of an elemental analyzer and a mass spectrometer (Reineking, Langel, & Schikowski, 1993). Isotopic signatures are expressed using the δ notation as δX (‰) = (R sample − R standard )/R standard × 1,000, with X representing the target isotope and R the ratio of heavy to light isotope ( 13 C/ 12 C and 15 N/ 14 N). For δ 15 N and δ 13 C analyses, N in atmospheric air and Vienna PD Belemnite served as standards respectively.

| Extraction and derivatization of AAs
For CSIA, dried samples were transferred to Pyrex culture tubes and flushed with N 2 gas, sealed and hydrolyzed in 6 N HCl at 110°C in a heating block for 20 hr (Larsen et al., 2013). After hydrolysis, lipophilic compounds were removed by adding n-hexane/DCM to the Pyrex tubes that were flushed shortly with N 2 gas and sealed before they were vortexed for 30 s. The aqueous phase was then filtered through a Pasteur pipette lined with glass wool that had been pretreated at 450°C. All samples were transferred into 4-ml dram vials before evaporating the samples to dryness under a steam of N 2 gas at 110°C in a heating block for 30 min. The samples were then stored at −18°C. To volatize the AAs, we followed the derivatization procedure of Corr, Berstan, and Evershed (2007), methylating the dried samples with acidified methanol and subsequently acetylating them with a mixture of acetic anhydride, trimethylamine, and acetone (1:2:5) to produce N-acetyl methyl ester derivatives. To reduce oxidation of AAs during derivatization, reaction vials were flushed and sealed with N 2 prior to the methylation and acetylation reactions.

| Compound-specific analysis of AAs
Amino acid derivatives were injected into a Thermo Trace GC coupled via a GP interface to a Delta Plus mass spectrometer (Thermo).
The GC was equipped with an Agilent J&W VF-35ms GC column (30 m × 0.32 mm × 1.00 µm). The temperature programme started with 80°C held for 1 min, increased by 20°C per minute to 135°C, then by 5°C per minute to 160°C and held for 3 min, then increased again by 8°C per minute to 300°C and held for 3 min. The injection temperature was 280°C and helium was used as carrier gas. The flow rate of helium was 2 ml/min. All samples were analysed in triplicate. To account for C added during derivatization and variability of isotope fractionation during analysis, pure AAs with known δ 13 C and δ 15 N values were also derivatized and analysed. Nor-leucine was used as internal reference.
The N isotopic composition of AAs in samples was expressed relative to atmospheric N by normalizing measured values (vs. reference gas) using scales derived from known δ 15 N values of the reference mixture.
The C isotopic composition was corrected for carbon added during derivatization following O'Brien, Fogel, and Boggs (2002) and was expressed relative to Vienna PD Belemnite.
Trophic fractionation of N isotopes (Δ 15 N C−D ) in AAs was calculated as the difference in δ 15 N values between consumer and diet (δ 15 N consumer − δ 15 N diet ), that is, between fungi and leaf litter, and between Diptera and fungi. Differences in Δ 15 N C-D values were analysed using ANOVA, with fungal functional groups (litter saprotrophic fungi, humus saprotrophic fungi and ECM fungi) and AAs as fixed factors. To test whether Δ 15 N C-D differed significantly from zero we conducted one sample t tests for each AA within fungal functional groups separately.
Amino acid-based trophic position of fungi and Diptera (TP CSIA ) was calculated using the following equation (Chikaraishi et al., 2009(Chikaraishi et al., , 2014: The equation takes into account the difference in δ 15 N values between Glu and Phe in the primary producer (β). Since β was similar among litter types, we used the mean value of 12.7 ± 1.4‰. The trophic discrimination factor (TDF Glu−Phe ) did not differ significantly between fungi and animals in controlled feeding studies (Steffan et al., 2015); we therefore adopted the proposed value of 7.6 ± 1.2‰ (Chikaraishi et al., 2009(Chikaraishi et al., , 2014 to calculate TP CSIA . When calculating TP CSIA , we propagated variance by using the differential solution of Equation 1 according to Blum, Popp, Drazen, Anela Choy, and Johnson (2013) and Ohkouchi et al. (2017). We used 'trophic isoclines' (Chikaraishi et al., 2014)  To test whether trophic position differed between saprotrophic fungi (litter and humus saprotrophic fungi pooled) and ECM fungi, we conducted an unpaired two sample t test assuming equal variances and using a confidence level of 0.95. Correlations between trophic positions as calculated by bulk stable isotope analyses and CSIA were tested using linear models for mean trophic positions of fungi. We tested for differences in the fractionation between leaf litter and fungi (Δ 13 C fungi−litter ) and between fungi and Diptera larvae (Δ 13 C Diptera−fungi ) in fungal functional groups using MANOVA and ANOVAs for single AAs. Δ 13 C values were calculated between individual pairs of leaf litter and fungi collected at the same site, and between Diptera larvae and the fungal sporocarps they were collected from, thereby accounting for spatial variability of isotopic signatures. For analyses of Δ 13 C, we grouped AAs into essential AAs that cannot be synthesized by animal consumers, and non-essential AAs that can be synthesized by animal consumers. This differentiation is not valid for fungi as they can also synthesize AAs essential to animal consumers, but it can be useful to identify AAs that are suitable for AA stable isotope fingerprinting (Larsen et al., 2009(Larsen et al., , 2013. We classified the following AAs as essential: isoleucine-Ile, leucine-Leu, methionine-Met, phenylalanine-Phe, threonine-Thr and valine-Val (Nation Sr., 2015;O'Brien et al., 2002); alanine-Ala, asparagine-Asp, glutamate-Glu, glycine-Gly, proline-Pro and tyrosine-Tyr were classified as non-essential. To test whether Δ 13 C fungi−litter and Δ 13 C Diptera−fungi differed significantly from zero, we conducted one sample t tests for each AA within fungal functional groups separately. To predict the biosynthetic origin of AAs in Diptera larvae (the 'fingerprinting' approach), we performed linear discriminant function analysis (LDA) with δ 13 C values of the essential AAs Ile, Leu, Met, Phe, Thr and Val, using litter and humus saprotrophic fungi, ECM fungi, and leaf litter as classifier variables. All analyses were performed in R (version 3.5.1 'Feather Spray').

| δ 15 N values of amino acids
saprotrophs, with the exception of Asp and Thr having higher and lower Δ 15 N fungi−litter in humus saprotrophic fungi compared to litter saprotrophic fungi, respectively (see Figure S1). Despite higher  (Table 1; Figure 2).
In Diptera, TP CSIA ranged between 2.3 and 3.7 (Table 1) trophic position lower compared to TP CSIA ; Table 1; Figure 3). The trophic position of ECM fungi and humus saprotrophic fungi was estimated slightly higher using bulk SIA than when using CSIA (Figure 3).
Trophic position of Diptera (Pegomya sp. and M. fungorum) was estimated 1.2 trophic position higher when using bulk SIA compared to CSIA (Table 1).

| δ 13 C values of amino acids
The fractionation of 13 C between leaf litter and fungi (Δ 13 C fungi−litter ) in AAs differed significantly between the different groups of fungi (ECM fungi, humus saprotrophic fungi and litter saprotrophic fungi; MANOVA; F 22,30 = 14.26, p < 0.0001; Figure 4). In saprotrophic fungi, the essential AA Met and Ile had a significantly higher Δ 13 C fungi−litter than in ECM fungi (ANOVA; F 2,25 = 41.88, p < 0.0001 and F 2,25 = 3.92, p = 0.033, respectively). The Δ 13 C fungi−litter of the non-essential AA Glu was significantly higher in humus saprotrophic fungi compared to litter saprotrophic fungi and ECM fungi (ANOVA;

F I G U R E 4
Trophic fractionation between leaf litter and fungi (Δ 13 C) in essential and non-essential amino acids. Fractionation differs between saprotrophic fungi decomposing litter (yellow triangles) and humus (orange circles), and ectomycorrhizal fungi (blue squares). Significant differences between fungal groups are indicated by asterisks (***p < 0.0001, *p < 0.05)  non-essential AA Asp did not differ significantly from zero in ectomycorrhizal and humus saprotrophic fungi, and Pro and Ala in ectomycorrhizal and litter saprotrophic fungi (for mean and SD of Δ 13 C between leaf litter and fungi, and p-values for one sample t tests refer to Table S1).
Trophic fractionation of 13 C between fungi and Diptera larvae (Δ 13 C Diptera−fungi ) differed between Diptera feeding on ECM fungi and on saprotrophic fungi (ANOVA; F 1,76 = 54.93, p < 0.0001), and also between AAs (ANOVA; F 10,76 = 3.14, p < 0.001). While Δ 13 C Diptera−fungi was close to zero in most essential AAs of Diptera larvae feeding on ECM fungi, fractionation was larger in those feeding on saprotrophic fungi ( Figure 5). In M. quadrimaculata, only Δ 13 C Diptera−fungi of the AA Met and Ser differed significantly from zero; in M. fungorum Δ 13 C Diptera−fungi of the non-essential AAs Ala and Asp, and of the essential AAs Leu and Val differed significantly from zero. For Mycetophilidae that fed on humus saprotrophic fungi there was only one measurement in which Δ 13 C Diptera−fungi of the AAs Ile and Met was close to zero. In Pegomya sp. Δ 13 C Diptera−fungi differed significantly from zero in the essential AAs Leu and Phe as well as in the non-essential AAs Ala and Asp ( Figure 5; Table S2).
Linear discriminant function analysis (LDA) of essential AA δ 13 C values of leaf litter and fungi clearly separated fungi from F I G U R E 5 Trophic fractionation (Δ 13 C) in essential and non-essential amino acids between fungi and Diptera feeding (a) on ECM fungi (Mycetophila fungorum and Metalimnobia quadrimaculata) and (b) on saprotrophic fungi (unidentified individuals of Mycetophilidae and Pegomya sp.). Note that trophic fractionation is higher in Diptera species feeding on saprotrophic fungi than in those feeding on ECM fungi. Significant differences from zero (p < 0.05) are indicated by asterisks  Table 2; F 6,88 = 50.68, p < 0.0001). Separation on LD1 was mainly due to Leu and Phe, whereas separation on LD2 was due to Leu, Met and Ile (for group means and coefficients of linear discriminants see Table S3). Diptera feeding on saprotrophic fungi were separated from those feeding on ECM fungi on LD1, but not on LD2. Despite higher offsets in Δ 13 C Diptera−fungi in Diptera feeding on saprotrophic fungi, they were positioned closer to their resource in LDA than Diptera feeding on ECM fungi. The latter were closer to saprotrophic than to ECM fungi on LD2 (Figure 6).

| D ISCUSS I ON
From a food web perspective, the distinction between ECM and saprotrophic fungi and their role in the nutrition of plants, soil micro-organisms and soil animals, has been a long-standing question.
Here we investigated the trophic position of these fungal functional groups by measuring bulk isotopic and AA compound-specific δ 15 N values of fungal sporocarps. In addition, to trace differential C utilization, we measured compound-specific δ 13 C values of AAs in leaf litter, saprotrophic and ECM fungi. We compared fungal signatures to those of associated Diptera larvae to explore whether their bulk F I G U R E 6 Scatterplot of the first two linear discriminants in linear discriminant function analysis (LDA) of δ 13 C values in essential amino acids (Ile, Leu, Met, Phe, Thr, Val). Fungi (litter saprotrophic fungi in yellow, humus saprotrophic fungi in orange, ectomycorrhizal fungi in blue) and leaf litter (in green) were used as classifiers to predict biosynthetic origin of amino acids in fungal feeding Diptera larvae. Diptera feeding on ectomycorrhizal (in light blue) and saprotrophic fungi (in pink) are drawn with convex hulls LDA of fungi, leaf litter and Diptera Elipses encircle 75% of the data for visualization purposes.

Mahalanobis distances
Sap ( TA B L E 2 Squared Mahalanobis distances for linear discriminant function analysis between litter saprotrophic fungi (Sap (lit)), humus saprotrophic fungi (Sap (hum)), ectomycorrhizal fungi (Myc) and leaf litter (Litter) and compound-specific signatures reflect consumption of fungi belonging to different functional groups.
Differential fractionation of C isotopes in ECM and saprotrophic fungi at least in part is due to differences in substrate utilization, but differences in physiological processes may also contribute to it (Henn & Chapela, 2001). Saprotrophic fungi presumably utilize more 13 C-enriched components of plant remains, such as cellulose, or microbially processed material, whereas ECM fungi use simple sugars supplied by their host plants (Hobbie et al., 1999). Furthermore, differences in respiration-related 13 C discrimination between ECM and saprotrophic fungi may also contribute to the 'EM-SAP Divide' (Boström et al., 2008).
Differences in the fractionation of N may result from more complex mechanisms and potentially include taxon-specific N uptake, transfer of amines to plant hosts, varying fluxes between N processing pathways and differential N allocation between tissues (Henn & Chapela, 2001). Gebauer and Taylor (1999) found

| Delineation of the trophic position of fungi and Diptera
The trophic fractionation of N between leaf litter and fungi (Δ 15 N fungi−litter ) was significantly higher in ECM fungi compared to saprotrophic fungi in most AAs, presumably because they utilize N from more humified organic matter with higher δ 15 N values (Gebauer & Taylor, 1999). Likely, utilization of more humified organic matter is also responsible for the higher Δ 15 N fungi−litter in most AAs of humus saprotrophic fungi compared to litter saprotrophic fungi. Humification is associated with repeated microbial processing and concurrent fractionation of N isotopes (Hyodo et al., 2008).
Interestingly, despite higher fractionation between leaf litter and fungi in various AAs of ECM fungi compared to saprotrophic fungi, the trophic position of ECM fungi and saprotrophic fungi only differed marginally, contrasting results of bulk isotopic analyses, where saprotrophic fungi occupied a significantly lower trophic position.
In part, this was due to considerable depletion of Phe in litter saprotrophic fungi compared to litter. Similarly, springtails and oribatid mites feeding on leaf litter were considerably depleted in 15 N in Phe compared to leaf litter . Both litter saprotrophic fungi and primary decomposers feeding on leaf litter may preferentially utilize specific compounds/tissues having lower 15 N values than bulk litter material. By contrast, humus saprotrophic fungi and ECM fungi presumably utilize more decomposed material to acquire N from soil. The 15 N value of Phe is likely to be close to the utilized substrate, since due to its complex structure it is more efficient for fungi to incorporate it from the substrate than to synthesize it themselves.
Micro-organisms feeding on plant tissues do not differ from animal consumers with respect to AA 15 N fractionation (Steffan et al., 2015;Steffan & Dharampal, 2018). Thus, the trophic position of fungi as calculated from 15 N CSIA of AAs may be more accurate than that derived from bulk stable isotopes since it accounts for isotopic variation of utilized substrates, counterbalancing differences between functional groups of fungi. However, large depletion of Phe in fungi compared to leaf litter may also originate from de novo synthesis of this AA by fungi, or from partial degradation of Phe in leaf litter. Fungi with the ability to degrade structural compounds of plants use enzymes such as oxygenases and peroxidases that break down aromatic compounds including lignin, but also Phe (Fuchs, Boll, & Heider, 2011;Janusz et al., 2017). Lack of Phe in conjunction with availability of N from lignin degradation may lead to enhanced synthesis of this AA in some fungal species. Usually, mostly saprotrophic fungi are able to degrade lignin (Tanesaka, Masuda, & Kinugawa, 1993), however, ectomycorrhizal fungi of the genus Lactarius also possess polyphenol oxidases (Giltrap, 1982). Indeed, besides saprotrophic fungi, L. flexuosus also had considerably depleted values of δ 15 N Phe . Presumably, CSIA of 15 N can provide more detailed insights into physiological processes and specific utilization of resources than bulk isotope analysis. However, there are still potential pitfalls when interpreting δ 15 N signatures in fungi irrespective of whether bulk or CSIA signatures are considered. For instance, various fungi are capable of utilizing inorganic N from soil (Gebauer & Taylor, 1999;Leake, Donnelly, & Boddy, 2002), possibly leading to differences in δ 15 N values; and there may be bidirectional transport of N compounds (glutamine, ammonium) between fungi and host plants in mycorrhizal fungi (Chalot & Brun, 1998;Martin & Botton, 1993).
TP CSIA for the Diptera Pegomya sp. and M. fungorum matched very well the expected trophic position, that is, they were positioned ~1 trophic level above the respective fungi they fed on. This contrasts the trophic positions of the same species as calculated from bulk 15 N values that were considerably higher than expected (see above), suggesting that TP CSIA is more accurate than TP bulk for these species. By contrast, due to considerable 15 N enrichment of Phe, TP CSIA of M. quadrimaculata and an unidentified individual of the family Mycetophilidae were markedly lower than expected. Pronounced enrichment of Phe (6-7‰) in the latter may be due to selective utilization of only certain fungal compounds such as protein as discussed above.

| δ 13 C values of amino acids
Saprotrophic fungi differed significantly in essential AA δ 13 C values from ECM fungi, as shown in LDA, which separated fungi from leaf litter along the first and ECM from saprotrophic fungi along the second LDA axis. Consequently, CSIA of 13 C in AAs has the potential to distinguish not only between plants and fungi, but also between the two ecological groups of fungi, that is, ECM and saprotrophic fungi. This may also allow to separate animal consumers of ECM and saprotrophic fungi using the fingerprinting approach (Larsen et al., 2009(Larsen et al., , 2013, opening new perspectives for soil animal food web analyses, as it is still an open question whether mycorrhizal fungi are consumed in significant amounts by fungal feeding soil invertebrates (Bluhm et al., 2019;Jonas, Wilson, White, & Joern, 2007;Pollierer, Langel, Körner, Maraun, & Scheu, 2007;Potapov & Tiunov, 2015).
Fractionation of 13 C between leaf litter and fungi also differed between ECM and saprotrophic fungi. In general, AA δ 13 C values of ECM fungi were closer to leaf litter than AA δ 13 C values of saprotrophic fungi. Closer AA δ 13 C values to leaf litter in ECM fungi presumably are due to the fact that ECM fungi use simple C compounds from their host trees, that is, sugars allocated to roots, whereas saprotrophic fungi predominantly use more complex litter compounds such as cellulose which are enriched in 13 C (Chen et al., 2019;Gleixner et al., 1993;Henn & Chapela, 2001;Kohzu et al., 1999). Δ 13 C fungi−litter did not differ significantly from zero in Phe in humus saprotrophic and ECM fungi, suggesting that in these fungi this complex AA is preferentially taken up from the substrate rather than synthesized de novo. In litter saprotrophic fungi, Phe in trend was slightly depleted in 13 C compared to litter, potentially due to de novo synthesis as also suggested by depleted δ 15 N values of Phe. Fractionation of Ile was also close to zero in humus saprotrophic fungi and ECM fungi. This is in line with the findings of Abraham and Hesse (2003) who suggested that isotope ratios of Ile are close to the used substrate in different fungal species. By contrast, the fractionation of Met differed considerably between saprotrophic and ECM fungi, being ~10‰ enriched in saprotrophic fungi, but close to zero in ECM fungi. Abraham and Hesse (2003) showed that C isotope ratios of AAs in fungi followed closely those of the substrates they utilized, with specific patterns depending on the used substrate. Similar δ 13 C values of Met in ECM fungi and leaf litter suggest that ECM fungi mainly used freshly fixed plant carbon to synthesize Met, whereas saprotrophic fungi used a different, 13 C-enriched source for synthesis of Met, potentially derived from 13 C-enriched components of leaf litter such as cellulose or starch (Pollierer et al., 2009). However, Abraham and Hesse (2003) also found phylogenetic differences in isotopic offsets between fungal species and suggested these to be related to genetically determined pathways and enzymatic reactions that are more similar in more closely related species. Yet, as saprotrophic and mycorrhizal strategies are independent of phylogeny (James et al., 2006), and we found consistent differences in AA 13 C patterns, we propose that isotopic offsets in C between ECM and saprotrophic fungi are determined predominantly by ecology and not by phylogeny.
Diptera feeding on saprotrophic fungi had higher isotopic offsets between resource and consumer (Δ 13 C Diptera−fungi ) than Diptera feeding on ECM fungi. Significant fractionation even in essential AAs in Diptera may be caused by gut microbial supplementation as has been shown for cerambycid beetles (Ayayee et al., 2016) and enchytaeids . However, fractionation may also stem from the utilization of specific fungal tissues or compounds, as discussed above for 15 N values of AAs.
Despite higher trophic fractionation, the fingerprinting analysis positioned Diptera feeding on saprotrophic fungi close to their resource, whereas Diptera feeding on ECM fungi were shifted towards saprotrophic fungi on LD2. This was due to significant fractionation in the essential AAs Leu and Met in these species, with the latter AAs being mostly responsible for separation of fungal functional groups on LD2. By contrast, the Δ 13 C Diptera−fungi of the essential AA Ile, Phe and Thr did not differ significantly from zero in any of the investigated species, suggesting that these AAs may be used to trace the fungal diet in consumers. In addition, the essential AA Val only differed significantly from zero in one of the investigated species. Overall, the attempt to assign Diptera to their respective fungal diet using 13 C fingerprinting was not as straightforward as we assumed, presumably due to selective consumption and utilization of certain fungal tissue and/or compounds and gut microbial supplementation, resulting in differential values of Δ 13 C Diptera−fungi even in essential AAs. However, as many soil animals consume fungal hyphae, which presumably also differ in 13 C values of essential AAs (Larsen et al., 2009), their separation by 13 C fingerprinting is promising.

| CON CLUS IONS
Here for the first time we compared compound-specific stable isotope composition of N and C in AAs of saprotrophic and ECM fungi.
Determination of trophic position of fungi by δ 15 N CSIA of AAs accounts for isotopic variability of utilized substrates and therefore yielded more realistic results compared to bulk 15 N analysis, with converging trophic positions of ECM and saprotrophic fungi. Thereby, the approach may allow better understanding of the contribution of major groups of fungi to the flux of energy through soil food webs. Importantly, saprotrophic and ECM fungi could be separated by their AA δ 13 C values, opening the potential of separating not only functional groups of fungi, but also their consumers in fingerprinting approaches. This may help to resolve some long-standing questions on how plant C and N are channelled to higher order consumers in soil food webs. However, as consumers may selectively feed on different fungal tissues (caps, stipes, hyphae) or preferentially utilize certain compounds (chitin, protein), as presumably is the case in fungal feeding Diptera larvae, future studies need to explore tissue-/compound-specific isotopic differences in fungi and how they affect consumer isotopic AA signatures.

ACK N OWLED G EM ENTS
We thank Jens Dyckmans for CSIA of amino acids. We thank Dr A.V.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are deposited in the Dryad Digital Repository https://doi. org/10.5061/dryad.vhhmg qnrb .