Interspecific temporal and spatial differences in the acquisition of litter-derived nitrogen by ectomycorrhizal fungal assemblages


Author for correspondence:

Andrea Polle

Tel: +49 551 39 3480



  • The spatiotemporal dynamics of, and interspecific differences in, the acquisition of litter-derived nitrogen (N) by natural assemblages of ectomycorrhizal root tips are poorly understood.
  • Small cylindrical mesh bags containing 15N-labelled beech (Fagus sylvatica) leaf litter that permit hyphal but not root ingrowth were inserted vertically into the top soil layer of an old-growth beech forest. The lateral transfer of 15N into the circumjacent soil, roots, microbes and ectomycorrhizas was measured during an 18-month exposure period.
  • Ectomycorrhial fungi (EMF) showed large interspecific variation in the temporal pattern and extent of 15N accumulation. Initially, when N was mainly available from the leachate, microbes were more efficient at N immobilization than the majority of EMF, but distinct fungal species also showed significant 15N accumulation. During later phases, the enrichment of 15N in Tomentella badia was higher than in microbes and other EMF species. Roots and soil accumulated 15N with a large delay compared with microbes and EMF.
  • Because approximately half of the studied fungal species had direct access to N from leaf litter and the remainder to N from leached compounds, we suggest that EMF diversity facilitates the N utilization of the host by capturing N originating from early-released solutes and late degradation products from a recalcitrant source.


In temperate forest ecosystems that are not markedly affected by anthropogenic emissions, nitrogen (N) is a major limiting nutrient (Vitousek et al., 2002). The turnover of litter accumulated on the forest floor potentially represents the main source of N for plant growth (Hart & Firestone, 1991). In addition to soil microbes, ectomycorrhizal fungi (EMF), which are associated with the root tips of the dominant tree species in these ecosystems, contribute to the degradation of leaf litter and the release of nutrients through the secretion of hydrolytic and oxidative enzymes (Näsholm et al., 2009). Depending on the environmental conditions, different EMF species exude different exoenzymes to make nutrients available (Abuzinadah & Read, 1986; Courty et al., 2005; Buée et al., 2007; Koide et al., 2008; Bödeker et al., 2009). Furthermore, the morphology of external hyphae emanating from the colonized root tips differs strongly among EMF species. Based on structure, abundance and lengths of the external mycelia, EMF have been classified as contact, short-distance, medium-distance, and long-distance exploration types, which, in addition to their hydrophilic and hydrophobic properties, may contribute differentially to host nutrient acquisition (Agerer, 2001). Although the potential of EMF for mobilization of nutrients from different sources and their ability to explore different soil zones are well known (Hobbie & Horton, 2007; Pritsch & Garbaye, 2011), the analysis of distinct functions of individual EMF species within their natural community is notoriously difficult. Stable isotope signatures, which are the result of discrimination of naturally occurring 15N/14N or 13C/12C in metabolic processes, have been employed to distinguish major N and carbon (C) sources of fungi (Hobbie & Högberg, 2012). EMF-ensheathed root tips generally show higher natural 15N abundance than nonmycorrhizal root tips because of fungal 15N enrichment (Zeller et al., 2007; Högberg et al., 2008; Tedersoo et al., 2012). In a detailed analysis, Tedersoo et al. (2012) showed that the interspecific variation of the natural 15N signature of ectomycorrhizas was related to fungal lineage, but correlations with exoenzyme activities or exploration types were not found (Tedersoo et al., 2012). Hobbie & Högberg (2012) linked interspecific variation in the mean 15N signatures of EMF to preferential N uptake from the mineral horizon or the organic layer. However, the implications of different EMF species, which co-occur in assemblages, for in situ N acquisition from the same resource still remain unclear.

Beech (Fagus sylvatica), a dominant tree species forming large monospecific forests in Central Europe (Ellenberg & Strutt, 2009), produces leaf litter that is decomposed very slowly compared with that of other deciduous tree species (Jacob et al., 2010). To trace the fate of N from this recalcitrant source, 15N-labelled beech litter was applied to the forest floor (Zeller et al., 2000). The label appeared in ectomycorrhizas beneath the 15N litter cover after 6 months and reached the leaves of the mature trees after 9 months (Zeller et al., 2000). Whether the EMF species, which ultimately mediated N transfer, have different N acquisition rates within their field assemblages is currently unclear.

In the present study, we investigated the accessibility of litter-derived N for a typical EMF assemblage associated with beech roots in an old-growth beech forest. The goal was to identify interspecific differences in N acquisition from leaf litter by analysing the temporal pattern of N accumulation in root tips colonized by different EMF species in comparison with N accumulation in fine roots, soil microbial biomass, and soil. With regard to functional differences among EMF, we hypothesized that ectomycorrhizas with intermediate- and long-distance hyphae would have direct access to N in the litter bags and therefore accumulate 15N more rapidly than short-distance or contact EMF. We expected, therefore, to find interspecific differences in 15N enrichment initially in EMF species. We furthermore hypothesized that these differences would disappear over time as short-distance and contact EMF would obtain increasing access to dispersed 15N. To test these hypotheses, mesh bags were filled with 15N-labelled beech leaf litter and deposited for up to 18 months in the top soil layer (0.1 m) of an old-growth beech forest. This layer contains the majority of fine roots (Meinen et al., 2009). The mesh width of the litter bags allowed ingrowth of fungal hyphae but prevented direct access of the roots. 15N enrichment was measured in soil, microbial biomass, fine roots and root tips colonized by different EMF species during 18 months. To distinguish between N uptake from litter bags and from the soil, half of the litter bags were removed after 14 months and the 15N enrichment in microbial biomass, ectomycorrhizas and roots in the presence and absence of the litter bags was measured after further incubation for 4 months. Thereby, ectomycorrhizas foraging in the litter bags and the surroundings were distinguished.

Materials and Methods

Site description

The study site was located in a 70–80-yr-old beech (Fagus sylvatica L.) stand in Tuttlingen Forest, in a low mountain range in southern Germany (longitude 8°45′E; latitude 47°59′N) at an altitude of 760–820 m above sea level. The climate is temperate with a mean annual air temperature of 6.6°C and a total annual precipitation of 856 mm. The soil profiles are characterized as Rendzic Leptosol derived from limestone and marls (WRB-classification, ISSS 1998). The organic layer comprises L and Of horizons of different depths (0–7 cm). The Ah horizon has pH values of c. 6.1 and a C : N ratio of 14.2. The mean pH of the forest floor is 5.6 and the C : N ratio is 26.7. The study site is exposed to an atmospheric N deposition rate of < 10 kg N ha−1 yr−1. Further details of the site, stand, and soil have been provided elsewhere (Holst et al., 2004; Dannenmann et al., 2007a,b).

Litter and litter bags

15N-labelled beech leaf litter was produced by Zeller et al. (1998). Leaves were collected in the autumn and dried at room temperature. The N concentration in the leaf litter was 1.134% with 1.137 atom% 15N enrichment. The litter bags (10 × 3 cm) were manufactured from nylon mesh (A. Hartenstein GmbH, Würzburg, Germany) with a mesh width of 50 μm, which allows penetration by fungal hyphae but prohibits roots from growing into the litter bag. Each bag was filled with 5.00 g of dry beech leaves and sealed.

Field study and sampling design

A rectangular plot (12.0 × 17.0 m) comprising nine beech trees was established on 14 April 2008. Four soil cores (0.03 m diameter and 0.1 m depth) were removed at a distance of c. 1 m from the stem of each tree and c. 0.5 m apart from each other. The samples of each tree were pooled and used for further analyses (= 9). A litter bag was inserted into each hole generated by the soil corer in tight contact with the adjacent soil. The bags were numbered and their positions were marked with flags to enable retrieval. Subsequent harvests were conducted to a depth of 0.1 m with a soil corer of 0.08 m diameter positioned to extract the soil core with the litter bag in its centre (Fig. 1). One sample per tree was collected after 6 months (14 October 2008; = 9), 14 months (19 June 2009; = 6), and 18 months of litter bag exposure (8 October 2009; = 6). During the 18-month exposure period in the unfenced field plots, some of the bags were lost, and therefore the sample numbers for each time-point varied between 6 and 9. To distinguish between direct and indirect effects of the labelled litter, after 14 months, seven litter bags were removed and the soil cores at the marked positions of these bags were harvested after a further 4 months. Unlabelled soil cores were taken from an adjacent site for the determination of 15N natural abundance (= 7).

Figure 1.

Scheme of sampling unit with the cavity, in which the litter bag was inserted. The soil sample representing a 2.5 cm-wide soil ring around the cavity was harvested for the analyses.

Sampling of soil and roots and identification of ectomycorrhizas and hyphae

The litter bags were removed from the soil cores, and aliquots of leaf litter were sampled for isotope analysis. The litter was dried at 50°C for 2 d to determine leaf mass loss. Soil samples of each soil core were collected and kept for isotope analysis. The soil water content was determined gravimetrically as the weight loss of oven-dried samples (105°C, until constant mass) compared with fresh soil and calculated as (mass of fresh soil − mass of dry soil) × 100/mass of dry soil.

The roots were carefully removed from the soil cores and washed with deionized water. The biomass of roots in each soil core was determined. The soil cores contained only fine roots (diameter < 2 mm). The roots were spread under a dissecting microscope (Leica M205 FA; Leica, Wetzlar, Germany) and the number of root tips per soil core was counted. EMF on vital root tips were morphologically categorized using a simplified procedure described by Agerer (1987–2006), applying the following keys: color, shape, branching pattern, mantle surface texture, hyphal morphology and abundance. All different morphotypes were photographed (Leica DFC 420 C). EMF were classified according to exploration type as contact, short-distance, intermediate-distance, and long-distance types (Agerer, 2001). The morphotypes were collected and stored at −20°C. EMF species sufficiently abundant for isotope analyses were identified by internal transcribed spacer (ITS) sequencing. For this purpose, on each sampling date, c. 10 ectomycorrhizas of each selected morphotype from different samples were pooled and identified through DNA sequencing. The sequence information was deposited in the National Center for Biotechnology Information (NCBI) databank, and the accession numbers and photographs are shown in Supporting Information Fig. S1. Methods for DNA extraction, PCR and sequencing are described in Table S1.

Mycelia that had grown within litter bags were harvested by suspending c. 0.80 g of the well-mixed litter bag contents in water under a dissecting microscope (Leica M205 FA). The floating hyphae were carefully collected using a forceps. The hyphae from four litter bags per harvest date were pooled. DNA extraction, PCR, cloning and sequencing of the mycelia were performed as described in Table S1. For each sample, 64 clones were analysed by PCR, yielding a total of 24 bands of different size. One representative of each band size was used for sequencing. Top BLAST matches are listed in Table S1.

Analysis of carbon, nitrogen and 15N and calculations

Vital ectomycorrhizas were collected under a dissecting microscope. Root tips ensheathed in a fungal mantle were excised at the last lateral root ramification. Samples of root tips colonized with distinct EMF species were freeze-dried and weighed (0.3–0.7 mg) and placed in tin capsules (4 × 6 mm). Because the mass of the root tips colonized by different EMF species ranged from 5.6 to 12.5 μg per root tip (Fig. S1), 30–60 root tips were pooled to obtain one sample of a given EMF species.

After EMF analysis, fine roots were cut into small pieces (c. 3 mm) and mixed. Subsamples were randomly collected for isotope analysis. The remaining roots were dried at 65°C for 2 d and the dry mass was determined.

For the analysis of total carbon, N, and 15N, samples of roots, soil, and litter were freeze-dried and milled (MM2; Retsch, Haan, Germany). Ectomycorrhizas were used without milling. The samples were weighed (S4; Sartorius, Göttingen, Germany) and placed in tin capsules (4 × 6 mm for biological material and 5 × 9 mm for soil; IVA Analysentechnik e.K., Meerbusch, Germany). The measurements were made with an isotope ratio mass spectrometer (IRMS Deltaplus; Thermo Finnigan Mat, Bremen, Germany) coupled to an elemental analyser (EA 1108; Fisons, Rodano, Italy).

The N isotope composition was expressed as δ15N (‰) against the standard of atmospheric nitrogen (air):

display math(Eqn 1)

R, the ratio of 15N to 14N isotopes in the sample and in the standard.)

Because the relative enrichment in 15N was measured for each EMF species separately, the mean δ15N (‰) for the total number of vital ectomycorrhizas was calculated as the sum of δ15N (‰) of the different EMF species (EMF1, EMF2, EMF3 … EMFn) weighed by the mean weight of the root tips and the relative abundance of the respective species. The following equation was used for each soil core:

display math(Eqn 2)

biomass (EMFi), the mean dry mass of a single root tip colonized by a given EMF species i; RAi, the relative abundance of root tips of vital EMF species i in the soil core; total biomass EMF, the sum of the mean dry mass of a single EMF root tip of species + the mean dry mass of a single EMF root tip of species + the mean dry mass of a single EMF root tip of species + … the mean dry mass of a single EMF root tip of species n.) Because rare species were not included, the calculated δ15N (‰)EMF represents c. 90% of the vital root tips colonized by EMF.

15N loss from the leaf litter was calculated by subtracting the 15N content (μg) of the litter bags at a given date from that of the initial 15N content of the litter bags. The amount of 15N in the leaf litter was calculated as follows:

display math(Eqn 3)

with 15N (atom%) = 15N × 100/(14N + 15N) and the biomass being the dry mass of the leaf litter in the litter bag at a given sampling date.

Nitrogen and 15N in soil microbes

Total N and 15N enrichment in soil microbial biomass were determined based on the disruption of microbial cells as a result of freeze-drying (Islam et al., 1997; Rumpel et al., 2001). The applicability of this method to determine microbial biomass N was tested for the investigated soil in an additional experiment, revealing that the treatments of chloroform fumigation, freeze-drying and the combination of freeze-drying and chloroform fumigation resulted in a similar increase in extractable N (Fig. S2). The chloroform fumigation extraction was conducted as described by Dannenmann et al. (2007b).

Both field-fresh and freeze-dried soil samples were subjected to extraction with 0.5 M K2SO4 (dry soil : solution ratio 1 : 5) as described by Dannenmann et al. (2009). Total N in the extracts was determined using a chemoluminescence detector (CLD-TOC) coupled to a C and N analyser (multi C/N 3100; Jena Analytics, Jena, Germany). The 15N enrichment in the total dissolved N compounds in the extracts was determined by applying the diffusion method of Brooks et al. (1989) as described by Wu et al. (2011). Briefly, all dissolved N compounds in the extract were quantitatively oxidized to nitrate via alkaline potassium persulfate oxidation. Subsequently, nitrate was quantitatively reduced to ammonium via the addition of Devarda's alloy, and an NaOH-induced pH increase facilitated the diffusion of ammonia on acid filter traps prepared for elemental analysis by isotope ratio mass spectrometry (EA-IRMS) (Dannenmann et al., 2009; Wu et al., 2011). Microbial biomass N and 15N enrichment were calculated from the difference between field-fresh and freeze-dried samples (Wu et al., 2011), without application of correction factors. Per definition microbial biomass includes the smallest living organisms such as bacteria, fungi, protozoa, algae, actinomycetes, and nematodes (Hoorman et al., 2011). The fraction of soil microbes also includes hyphae emanating from ectomycorrhizal mantles.

Statistical analysis

Statistical analysis was performed using Statgraphics Plus 3.0 (StatPoint Inc., St Louis, MO, USA). A natural-logarithm or square-root transformation was used to meet the requirement of normal distribution of data and variance homogeneity, when necessary. When data transformation did not meet these requirements, the nonparametric Mann–Whitney U-test was applied instead of analysis of variance (ANOVA). Means or medians were considered to be significantly different from each other if  0.05.


Ectomycorrhizal root tips and microbes accumulate litter-derived N more rapidly than roots

After exposure of beech litter bags for one growing season from April to October in a beech forest, the mass loss of beech litter was c. 20%, and the loss at the end of the second growing season was c. 46% (Fig. 2). The mass loss resulted in corresponding losses of N (not shown) and of 15N (Fig. 2). The N concentration (mean ± SE for all time-points: 11.8 ± 0.4 mg N g−1 dry mass) and the C to N ratio of the beech litter (mean for all time-points: 40.8 ± 0.9) remained relatively stable during the 18-month exposure period. The soil water content did not vary significantly at the different sampling dates (mean for all time-points: 32.6 ± 1.4%).

Figure 2.

Decrease in litter mass (open symbols) and in the amount of 15N (closed symbols) in the litter bags. The bags were filled with 5.00 g of beech (Fagus sylvatica) litter and exposed in the soil of a beech forest for 18 months. Data are means of five to seven samples (± SE). When the error bars are not visible, they are smaller than the symbols. Different letters indicate significant differences at  0.05.

We followed the appearance of 15N in soil cores up to a distance of 25 mm from to the litter bags. We observed clearly distinct time courses of tracer accumulation: the label was detected in both ectomycorrhizas and soil microbial biomass after 6 months; in bulk fine roots it appeared after 14 months and in soil it appeared after 18 months (Fig. 3). Initially, the 15N signature was stronger in microbes than in ectomycorrhizas (Fig. 3), whereas subsequently no significant differences between these fractions were found. At all time-points, the 15N enrichment in ectomycorrhizas and microbes was much higher than that of fine roots and soil (Fig. 3). The relative incorporation of 15N into roots in the soil cores varied between sampling dates from 0.03 to 0.1% of the 15N released from the litter bag.

Figure 3.

Increase in δ15N in ectomyorrhizal fungal root tips, roots, and soil adjacent to the 15N-labelled beech (Fagus sylvatica) leaf litter. A soil ring (width 2.5 cm) around the litter bags was collected and used for the analysis of 15N enrichment. Data are means of six to nine samples (± SE). When the error bars are not visible, they are smaller than the symbols. Closed triangles, microbes; open circles, ectomycorrhizas; closed circles, roots; open squares, soil. Red symbols not connected by a line at 18R indicate mean label in samples 4 months after the removal of the labelled litter bags. Different letters indicate significant differences at  0.05.

The removal of the litter bags in early summer of the second season (14 months) suppressed subsequent 15N enrichment in roots, whereas ectomycorrhizas and microbes still displayed a net increase in 15N compared with earlier time-points (Fig. 3).

Ectomycorrhizal fungal diversity and interspecific differences in the acquisition of litter-derived N

To determine whether there were differences among EMF species in N acquisition from litter, we determined the composition of the EMF assemblages and analysed the enrichment of 15N in the colonized root tips. All root tips in the cylindrical soil samples encasing the litter bags (to a depth of 0.1 m) were counted. The mean number of vital root tips was 1549 ± 162 and did not vary significantly during the 18-month time course of this study. The fraction of nonmycorrhizal root tips was negligible (< 1% of all tips). Fourteen different EMF species were detected based on morphological differences and ITS sequencing (Fig. S1). The abundance of different EMF species fluctuated between sampling dates; Cenococcum geophilum was the most frequent species, colonizing between 26 and 47% of the root tips (Table S2).

Seven of the 14 detected EMF taxa occurred on each of the four sampling dates and could therefore be used to determine the time course of 15N accumulation in distinct species. These species were classified by inspection of mantle properties and extramatrical hyphal morphology according to the following soil exploration types: Russula cuprea, contact; Humaria sp., C. geophilum and Sebacina sp., short-distance; Tomentella badia, short- to medium-distance; Cortinarius sp., medium-distance; and Boletus pruinatus, long-distance exploration type (Fig. S1). Altogether, these seven fungal species colonized c. 90% of the root tips (Table S2). Three of these species were also detected in mycelia from litter bags: Cortinarius sp., Sebacina sp., and Tomentella sp. (Table S1).

In the absence of labelled leaf litter, the ectomycorrhizas with different fungal species did not show significant differences in the abundance of natural 15N compared with fine roots, except for Cortinarius sp., in which δ15N was enriched compared with the other EMF species (Table S3). The 15N signature of soil was intermediate between that of the majority of EMF species and that of fine roots (Table S3).

After exposure to labelled litter, all EMF species accumulated 15N above their natural signature (Fig. 4). However, the extent and time courses of accumulation varied considerably. After 6 months of exposure to labelled litter, 15N was significantly increased in all EMF root tips, except in those colonized by B. pruinatus (Fig. 4). Although mean microbial δ15N enrichment after 6 months of exposure was significantly higher than mean ectomycorrhizal δ15N enrichment (Fig. 3), the 15N enrichment in root tips associated with some EMF taxa (Tbadia, R. cuprea and B. pruinatus) was close to that found in the soil fraction of microbial biomass at this time-point (Fig. 4, inset table).

Figure 4.

Temporal variation of δ15N in root tips colonized by different ectomycorrhial fungal (EMF) species. Box plots represent two to eight samples (n = 2 for Boletus pruinatus, Sebacina sp. and Russula cupraea after 6 months and for Humaria hemisphaerica after 18 months in samples in the absence of the litter bag). The box indicates the 75% range of the data, with the horizontal line indicating the median, the square the mean and whiskers the minimum and maximum. Different letters in box plots indicate significant differences at  0.05 for temporal variation of 15N enrichment. 18R refers to data for samples analysed 4 months after the removal of the labelled litter bags. Table inset: different letters in columns indicate significant differences at  0.05 (log-transformed data) for 15N enrichment among EMF species colonizing the root tips and microbes at a given time-point. Tb, T. badia; Co, Cortinarius sp.; Bp, B. pruinatus; Hu, Humaria sp.; Cg, C. geophilum; Se, Sebacina sp.; Rc, R. cuprea; Mic. soil microbes; UEM, uncultured ectomycorrhiza.

In the following season (14 months of exposure), the root tips colonized by Cortinarius sp. displayed the highest 15N enrichments and those of the two ascomycetes Humaria sp. and C. geophilum the lowest (Fig. 4). Overall, at 14 months of exposure, the 15N enrichment of ectomycorrhizas formed with different fungal taxa was not different from that of soil microbial biomass (Fig. 4, inset table).

All EMF showed a strong increment in δ15N at the end of the second season (18 months of exposure; Fig. 4). However, the 15N acquisition of T. badia was much higher than that of the other EMF species (Fig. 4, inset table). In addition to T. badia, Cortinarius sp., B. pruinatus and Humaria sp. showed greater 15N enrichment after 18 months of exposure than the soil-localized microbes (Fig. 4, inset table). Russula cuprea exhibited the lowest δ15N accumulation of all analysed EMF taxa (Fig. 4).

After 14 months of exposure to 15N, half of the litter bags were removed to distinguish between 15N uptake from the bags and from the surrounding environment. Four months after litter bag removal (time-point 18R in Fig. 4), EMF species with different behaviours in response to litter bag removal could be distinguished (Fig. 4): decreased 15N accumulation was found in EMF of the medium- and long-distance exploration types (T. badia, Cortinarius sp. and B. pruinatus) and increased 15N signatures, similar to those present in samples with continued litter bag exposure (Fig. 4), were detected in the short-distance types (Humaria sp., C. geophilum and Sebacina sp.). This result indicates spatial differences in N foraging. Cortinarius sp. and Tomentella sp., whose hyphae were detected within the bags (Table S1), must have had direct access to the litter bags, whereas the other species obtained N from their surroundings.

To determine whether the differences in 15N accumulation after 18 months of exposure to litter bags were a result of differences in fungal N demand resulting from differences in mantle thickness, we measured the N concentrations of root tips colonized by different EMF species and determined their biomass (Table S3). With these data, assuming that the N concentration of the plant portion in the mycorrhiza was similar to that of fine roots, we were able to estimate the amount of N in the different fungal partners of root tips (Table S3). We did not find a relationship between the 15N signature and the N concentrations of the roots tips, or between the 15N signature and the estimated amount of fungal N in the ectomycorrhizas (Fig. S3A,B). Therefore, differences in 15N accumulation in ectomycorrhizas do not reflect differences in N concentrations in the fungal biomass or in the root tips but rather indicate interspecific differences in N acquisition from degrading litter.


Accessibility of litter-derived N for microbes, ectomycorrhizas and plants

Our study shows a clear hierarchy for mean 15N tracer accumulation, first in microbes and ectomycorrhizas, then in roots and finally in soil. An important finding was that EMF species showed large interspecific variation in 15N accumulation, with some fungal taxa apparently being similarly efficient to soil microbes in N acquisition at an early stage.

The appearance of the label in the investigated fractions is the result of litter decomposition, translocation and 15N uptake. The decay rates in the litter bags were similar to those reported for litter exposed on the floor of temperate beech forests (Zeller et al., 2000, 2001; Stoelken et al., 2010). The initial mass loss of litter is the result of leaching of soluble compounds and subsequently of the degradation of recalcitrant biopolymers (Berg & McClaugherty, 2008). The transport of these compounds downwards to lower soil layers may result in a nutrient gradient with stratification of EMF species and saprotrophes (Dickie et al., 2002; Hobbie & Horton, 2007; Lindahl et al., 2007). In contrast to previous studies, the ectomycorrhizas, soil microbes and roots analysed here were not located at increasing distance beneath the decomposing labelled leaf litter (Zeller et al., 2000, 2001), but accumulated 15N by lateral N transfer from the adjacent litter bags. The mechanisms are not entirely clear, but probably involved dispersion with precipitation (Flühler et al., 1996) and hyphal transport.

Microbes incorporated the released N and showed peaks for the 15N label and microbial biomass (not shown) in autumn of the first and second years of exposure to leaf litter. This result supports earlier observations that microbial N cycling in beech forests shows seasonal patterns, with peaks of temporal N immobilization in the autumn and winter, whereas increased mineralization, accompanied by lower microbial N uptake, occurs preferentially in the summer months (Zechmeister-Boltenstern et al., 2002; Kaiser et al., 2010, 2011). It has been suggested that the transient immobilization of N in microbial biomass in autumn is a mechanism for nutrient conservation preventing N loss from the ecosystem (Kaiser et al., 2011). The results of our study support this hypothesis and indicate that EMF contribute to this ecological function but probably at a slower pace than soil-localized microbial fractions.

We expected that removal of the litter bags would result in net decreases in δ15N in all compartments because the 15N source was lacking and net efflux of N occurring. However, this decrease was not observed, underlining the importance of microbes and EMF for N retention. The 15N signatures in ectomycorrhizas and in microbes, surprisingly, even continued to increase, although to a much lesser extent than in the neighbourhood of the litter bags. One explanation for the observed increase is recycling of 15N from their own hyphae, transport from other storage sites or degradation of dead roots and their attached EMF. There is now emerging evidence that N from root necromass, which accounts for c. 50% of the root tips in this forest, is more readily available for beech nutrition than N from leaf litter (Pena et al., 2010; Guo et al., 2013). Furthermore, N from decomposing EMF-associated root tips would also have been available to the soil microbes via the same mechanisms as N from degrading leaf litter, leading to the observed 15N enrichment in the absence of litter bags.

Functional diversity of ectomycorrhizal species with respect to N acquisition from soil or leaf litter

Functional differences in fungal communities have mainly been studied through analysis of the natural abundance of stable isotopes (Lilleskov et al., 2002; Taylor et al., 2003; Hobbie & Hobbie, 2006; Tedersoo et al., 2012), which have the disadvantage that temporal changes can hardly be resolved, or by analysis of exo-enzyme activity profiles, which provide information on the potential for resource mobilization (Buée et al., 2005, 2007; Courty et al., 2010; Diedhiou et al., 2010; Tedersoo et al., 2012) but not on the actual nutrient acquisition. Our study is unique in providing direct evidence for interspecific differences in N acquisition in situ among co-occurring EMF species, resolving temporal patterns and revealing access to spatially distinct N sources.

The fungal assemblage of our study is typical for temperate beech forests (Lang et al., 2011) and was previously found in this forest (Pena et al., 2010). By removal of the 15N source we were able to distinguish between EMF that have access to N from sources other than leaf litter in their surroundings (Humaria sp., Sebacina sp. and C. geophilum) and those that must have received 15N by transport directly from the litter bags (T. badia and Cortinarius sp.). Indeed, the EMF species identified in each of these two categories by our labelling approach also correspond to morphologically distinct short- and medium-distance exploration types, respectively (Agerer, 2001). This finding implies that it will be advantageous for the host to associate with different fungal species to increase access to spatially differentiated N sources. This notion is also supported by the observation that increasing EMF diversity of beech roots was correlated with a significant decline in N in neighbouring roots of ash (Fraxinus excelsior Linn.), a species associated with arbuscular mycorrhizas (Lang & Polle, 2011). Those competitive advantages for the tree may stabilize diverse EMF communities on their host roots.

We also identified marked differences in temporal and quantitative changes in N accumulation among the fungal species. In contrast to our expectation, the long-distance exploration type B. pruinatus did not have primary access to N from leaf litter. As B. pruinatus has hydrophobic rhizomorphs, it may not be able to take up hydrophilic solutes (Hobbie & Agerer, 2010). It has been suggested that investment in long-distance transport by hydrophobic rhizomorphs to access dispersed resources would be profitable in nutrient-limited ecosystems (Hobbie & Agerer, 2010). Our results support this hypothesis, because significant 15N accumulation was observed with a long delay compared with other fungal exploration types.

All other studied EMF, including the contact type R. cuprea, exhibited significant 15N accumulation after 6 months, when a strong N signal was also present in the soil microbes, but no enrichment was yet found in roots. Russula cuprea has been described as a ‘protein fungus’ that has the capacity to mobilize organic N directly from proteins (Lilleskov et al., 2002). The lack of emanating hyphae indicates that N must have been absorbed from its immediate vicinity, and the extent of 15N enrichment in R. cuprea suggests a high competitiveness similar to that of soil microbes.

Among the short-distance exploration types, C. geophilum showed the lowest 15N accumulation, suggesting differences in N demand or N metabolism in comparison with Humaria and Sebacina. At our study site, the abundance of this EMF species on beech root tips was correlated with glutamine uptake (Dannenmann et al., 2009). Although C. geophilum apparently has the capacity to mobilize and utilize organic N (Abuzinadah & Read, 1986; Courty et al., 2005; Diedhiou et al., 2010), our study shows that its ability to accumulate litter-derived N is apparently not great in comparison with other EMF species.

Ectomycorrhizas of T. badia have been classified as being of the short- to medium-distance exploration type (Agerer, 2001). Tomentelloid ectomycorrhizas are colonizers of wood debris (Tedersoo et al., 2003) and have a high capacity to produce enzymes involved in litter degradation (Koljalg et al., 2000; Buée et al., 2005, 2007). In our study, T. badia ectomycorrhizas revealed the highest degree of 15N enrichment and were strongly responsive to litter bag removal. This result demonstrates their direct access to N from the litter and thus provides experimental evidence that T. badia ectomycorrhizas are strongly involved in N recapture from degrading leaves.

The Cortinarius genus comprises EMF of the medium-distance exploration type forming fan-like densely emanating rhizomorphs with hairy surfaces (Agerer, 2001). Cortinarius fungi live typically in organic soil and use protein and amino acids as their N source (Lilleskov et al., 2002; Cullings et al., 2003; Treseder, 2008). Here, Cortinarius sp. was the only analysed EMF species that showed enrichment of natural 15N in comparison with other EMF species, which is an indication of preferential use of organic N (Lilleskov et al., 2002). Our results clearly show that Cortinarius sp. captured N directly from degrading leaf litter. It is also interesting that Cortinarius sp. is among the most sensitive species to increased N deposition in forests (Brandrud & Timmermann, 1998). In the Tuttlingen beech forest, N deposition was moderate (Dannenmann et al., 2007b). The presence of Cortinarius sp. suggests that the threshold for N excess was not exceeded (Lilleskov et al., 2011).

In conclusion, our study shows strong interspecific variation in the ability of different fungal species associated with root tips to acquire N from degrading leaf litter. The extent of N enrichment was a species-specific feature and not related to differences in N concentration or the thickness of the fungal mantle. Notably, T. badia, a representative of the tomenteloids, was greatly superior to the bulk soil microbial fraction in N acquisition. However, T. badia was slow to realize its full potential in terms of N acquisition, which seemed to require colonization of the litter. Initially, when N was mainly available from the leachate, soil microbes were more efficient at N acquisition, but some fungal species, among them the contact type R. cuprea, also showed significant accumulation ability. The finding that approximately half of the fungal species had access to N directly from leaf litter and the remainder to N from soluble compounds in their immediate vicinity highlights the importance of EMF species diversity on root tips for the acquisition of both easily released ‘early’ and recalcitrant ‘late’ N as nutrient sources for beech.


We are grateful to T. Klein for help with the collection of the samples, to M. Fastenrath for help with morphotyping, to J. Dyckmans (Kompetenzzentrum für Stabile Isotope, University of Göttingen) and R. Meier (Center of Stable Isotopes of IMK-IFU) for 15N analysis and to the German Science Foundation for funding our work (Po362/17-1, Po362/17-2 and 459 Da1217/2-1).