Stable isotope natural abundances of fungal hyphae extracted from the roots of arbuscular mycorrhizal mycoheterotrophs and rhizoctonia-associated orchids

Since the first discovery of unique carbon (C) and nitrogen (N) isotope signatures in fungal fruiting bodies (Gebauer & Dietrich, 1993; Gleixner et al., 1993), natural abundances of stable isotopes have been extensively used to identify the nutritional dynamics of fungi (Mayor et al., 2009). Assigning ecological roles of fungi is essential to determine the role of individual taxa in nutrient cycling and forest ecology. The use of isotope natural abundances in forest ecosystems has been crucial in distinguishing fungi with two main modes of life: ectomycorrhizal and saprotrophic fungi (Henn & Chapela, 2001). Within saprotrophic fungi, isotope natural abundances further allow the identification of the substrates used (Kohzu et al., 1999). Dual isotope analyses of the δ¹³C and δ¹⁵N values consistently indicate a differentiation in isotopic signatures between ectomycorrhizal and saprotrophic fungi within and among ecosystems (Henn & Chapela, 2001; Taylor et al., 2003; Trudell et al., 2004; Mayor et al., 2009). These signatures have been shown to reflect the ecophysiology of fungi and demonstrate that fungi that can utilize organic nitrogen exhibit higher δ¹⁵N than those fungi restricted to mineral nitrogen sources (Gebauer & Taylor, 1999; Lilleskov et al., 2002). Still, the ability to distinguish fungal nutritional modes has been long restricted to fungi that produce macroscopic sporocarps, such as mushrooms, due to their large mass which allows for physical measurements. Thus, for many fungi, particularly those associated with plant roots that do not form evident fruiting bodies, isotope natural abundances of fungal hyphae are scarce.

Besides ectomycorrhizal fungi, isotope natural abundances are known for sporocarp-forming ericoid (e.g. Hobbie & Hogberg, 2012) and orchid-associated nonrhizoctonia saprotrophic fungi (e.g. Ogura-Tsujita et al., 2009). Yet, values of δ¹³C and δ¹⁵N are poorly known for arbuscular mycorrhizal fungi (but see e.g. Courty et al., 2011; Suetsugu et al., 2020, for isotope values of fungal spores), and the orchid-associated fungi known as ‘rhizoctonia’ in natural conditions. Recently, Klink et al. (2020) obtained the δ¹³C and δ¹⁵N of arbuscular mycorrhizal hyphae isolated from roots of a grass and a legume, inoculated in experimental conditions, thereby providing an efficient method to extract hyphae from roots. Using this method with a few modifications, here, we measured the isotope natural abundances δ¹³C and δ¹⁵N of naturally occurring arbuscular mycorrhizal (Fig. 1a–c) and orchid-associated hyphae (Fig. 1d–f) directly from roots (see Supporting Information Methods S1). To obtain hyphae of arbuscular mycorrhizal fungi, we selected two species of fully mycoheterotrophic plants: Thismia megalongensis C. A. Hunt, G. Steenbee. & V. Merckx and Sciaphila megastyla Fukuy. & T. Suzuki. Mycoheterotrophs are achlorophyllous plants that obtain carbon from their associated fungal partners (Leake, 1994; Merckx, 2013). Species in the plant genus Thismia have been demonstrated to be highly specialized on narrow lineages of Glomeromycotina fungi (Gomes et al., 2017; Merckx et al., 2017), while species of Sciaphila tend to associate with a wider phylogenetic diversity within the fungal subphylum (Merckx et al., 2012; Suetsugu & Okada, 2021). For fungi associated with orchid roots, we selected two chlorophyllous partially mycoheterotrophic orchid species, known to associate with rhizoctonia symbionts, Orchis militaris L. and Ophrys insectifera L., for which both isotope natural abundances and Sanger sequencing of the root-associated fungi have been performed previously (Schweiger et al., 2018). To be able to compare isotope values across sampling sites, the δ values of C and N stable isotope abundances were normalized by calculating enrichment factors (ε; see Methods S1).

The enrichment factors ε¹³C and ε¹⁵N were significantly different between the fungal hyphae, mycoheterotrophic and reference plants for both T. megalongensis and S. megastyla (Fig. 1g; Table 1). For both species, ε¹³C was not distinguishable between the mycoheterotrophs and respective fungal hyphae, while ε¹⁵N was significantly different between mycoheterotrophs and fungi for S. megastyla, and marginally significant for T. megalongensis (Fig. 1; Table 1). In relation to the reference plants, the fungi extracted from both mycoheterotrophic species were significantly enriched in ε¹³C, and fungi from S. megastyla were marginally significantly depleted in ε¹⁵N. Similarly, both mycoheterotrophic plants were enriched in ε¹³C although only significantly for T. megalongensis. This indicates that the ε¹³C of fungal hyphae drives the ¹³C enrichment of arbuscular mycorrhizal fully mycoheterotrophic plants, and there seems to be a difference in nitrogen source between T. megalongensis and S. megastyla-associated fungi. Each mycoheterotrophic plant species is associated with nonoverlapping fungal clades within the Glomeromycotina (Fig. 2a). Sciaphila megastyla harboured fungi belonging to the genera Dominikia, Kamienskia and two unidentified amplicon sequence variants, while the fungi in the roots of T. megalongensis belonged exclusively to the genus Rhizophagus, supporting a specialization on fungal interactions of different degrees between these plant lineages (Gomes et al., 2020; Suetsugu & Okada, 2021).

The enrichment factors ε¹³C and ε¹⁵N were generally significantly different between fungal hyphae, orchids and reference plants for both O. militaris and O. insectifera (Fig. 1h; Table 1). Both ε¹³C and ε¹⁵N were significantly different between orchid leaves and hyphae for O. militaris, while for O. insectifera, only ε¹³C was significantly higher in the hyphae in comparison with the plant tissue (Fig. 1; Table 1). In both orchid species, fungal hyphae were significantly enriched in ε¹³C, and in O. insectifera fungi were also enriched in ε¹⁵N in relation to the reference plants. The fungal hyphae extracted from the two orchid species were only weakly enriched in ε¹³C in comparison with reference plants and far less enriched in ¹³C than tissues of ectomycorrhizal fungi reported previously (Mayor et al., 2009). This observation is consistent with previous findings of absence of ¹³C enrichment in fully mycoheterotrophic protocorms of O. militaris, which were also associated with rhizoctonia fungi by Schweiger et al. (2018). Interestingly, in that study, protocorms of O. insectifera were somewhat enriched in ¹³C.

We detected most sequenced reads obtained from root pieces to belong to the fungal order Helotiales. Fungi in the genus Ilyonectria were also detected, concordant with previous observations of these orchid species collected at the same site (Schweiger et al., 2018). Both Helotiales and Ilyonectria were present in the roots of both orchid species and, as far as we know, have an unknown ecological function. Helotiales have also been detected in the species studied in Zahn et al. (2023). In addition, we detected rhizoctonia fungi belonging to the families Ceratobasidiaceae, Serendipitaceae and Thelephoraceae in the roots of O. insectifera, and to the families Ceratobasidiaceae and Thelephoraceae in the roots of O. militaris (Fig. 2b). One orchid individual of O. militaris presented a high relative abundance of Ceratobasidiaceae, and another of Thelephoraceae in their roots. We cannot exclude that Tulasnellaceae are underrepresented in our data influenced by the primers used (Vogt-Schilb et al., 2020), as these taxa have been shown to be present in O. insectifera roots (Schweiger et al., 2019). Besides rhizoctonia fungi, we also found fungi known to form ectomycorrhizas (according to FungalTraits; Põlme et al., 2020), such as Sebacina (Sebacinaceae), Amphinema (Atheliaceae), Hebeloma and Hymenogaster (Hymenogastraceae) in two O. insectifera individuals. In terms of isotope signatures, no apparent differences were observed between individual samples where rhizoctonia fungi are present and those where Helotiales are predominant, and neither in relation to the plant material between specimens. Yet, a larger sample size would be needed to properly evaluate this.

While ε¹³C values of hyphae are within the same range as found for the respective plant tissues of the AM mycoheterotrophic plants, as expected, ε¹⁵N values of the hyphae were considerably lower than ε¹⁵N of the respective plant tissues. This relative ¹⁵N depletion of hyphae in comparison with plant tissue was also observed for the two orchid species. One could wonder whether this depletion is either due to potential loss of hyphal content during extraction considering that nitrogen in chitin is depleted in ¹⁵N by c. 10‰ in comparison with fungal protein (Taylor et al., 1997; Hobbie & Hogberg, 2012) or due to a selective transport of ¹⁵N-enriched protein-derived compounds from fungal to plant tissues. Similarly, Zahn et al. (2023) show an equal depletion in ¹⁵N of hyphae extracted from two rhizoctonia-associated orchid species and for identically extracted hyphae from ectomycorrhiza-associated orchid roots an even larger depletion in ¹⁵N in relation to orchid leaves.

In addition, the N concentrations of the extracted fungal hyphae of both T. megalongensis (2.25 ± 0.53 mmol g_dw⁻¹) and S. megastyla (2.18 ± 0.42 mmol g_dw⁻¹) were not distinguishable from those of the mycoheterotrophic plant tissues (1.95 ± 0.28 and 1.88 ± 0.45 mmol g_dw⁻¹ respectively), in congruence with Klink et al. (2020), while reference plants presented lower N concentrations (1.36 ± 0.44 and 1.17 ± 0.19 mmol g_dw⁻¹ for each set respectively) in relation to both fungal hyphae (T. megalongensis: Z = 2.772, P = 0.008, and S. megastyla: Z = 3.231, P = 0.002) and mycoheterotrophic plants (T. megalongensis: Z = 2.140, P = 0.032 and S. megastyla: Z = 2.710, P = 0.007). The N concentrations between fungal hyphae (1.19 ± 0.23 mmol g_dw⁻¹ for O. militaris and 1.64 ± 0.17 mmol g_dw⁻¹ for O. insectifera), orchids (1.74 ± 0.07 and 2.19 ± 0.17 mmol g_dw⁻¹ respectively) and reference plants (1.62 ± 0.76 and 1.79 ± 0.79 mmol g_dw⁻¹ for each set respectively) were not statistically different for both orchid species. In Zahn et al. (2023), the fungal hyphae extracted from rhizoctonia-associated orchids were also nondistinguishable from reference plants, while for one species (Anoectochilus sandvicensis), fungal hyphae had significantly lower N concentration than the orchid leaves.

The arbuscular mycorrhizal diversity in the roots of the mycoheterotrophic plant species did not overlap between T. megalongensis and S. megastyla, and the fungal enrichment in ε¹⁵N was variable between plant species. The association with different fungal genera, in addition to local soil nitrogen availability, could have contributed to the differences in ε¹⁵N between species. Further studies are required to assess the source of variation and generality of isotope values among arbuscular mycorrhizal fungi.

In the orchid-associated fungi, the fungal composition was variable between individual specimens, yet without reflection on the isotopic values of the extracted hyphae. The absence of differences in fungal isotopic values may indicate an artefact on the integration of both techniques. The apparent dominance of specific fungal groups in the roots could reflect spatial segregation of fungi, as a small piece of root was used for sequencing, while for the hyphal extraction, the remainder of the root system was used. Furthermore, ectomycorrhizal fungi were detected in two individuals of O. insectifera, while rhizoctonia fungi were detected in three individuals. The presence of ectomycorrhizal fungi in the roots of some rhizoctonia-associated orchids is commonly reported in the literature (e.g. Jacquemyn et al., 2021), yet it remains to be demonstrated whether these fungi indeed establish a mycorrhizal symbiosis with the orchid, in a sporadic or constant way during the orchid development, or represent endophytic fungi as it has been shown in typical nonmycorrhizal hosts (Schneider-Maunoury et al., 2020). Nevertheless, we cannot exclude those ectomycorrhizal fungi found in the roots of O. insectifera to be responsible for the slight enrichment in ¹³C and ¹⁵N of the hyphae extracted from O. insectifera in comparison with O. militaris. However, these isotopic differences are rather small and are not seen in the leaves of these two species, that is there appears to be no major plant matter gain from these ectomycorrhizal fungi. In addition, our results reveal that sporadic appearance of ectomycorrhizal fungi in orchids hitherto classified as rhizoctonia-associated does obviously not affect their isotope signature. Zahn et al. (2023) present further isotope signatures and diversity of root-associated fungi of orchids associated with ectomycorrhizal fungi.

The assessment of fungal diversity often comprises a qualitative snapshot of a fraction of the root system, and although different fungal species or guilds may contribute differently to nutrient uptake at multiple occasions, we still lack a solid framework to quantify the contribution of each of these fungi to fungal–plant matter exchange. By contrast, isotopic abundance data are a temporal and spatial integrator (Dawson et al., 2002) over all fungal–plant matter exchange processes without providing direct information about the role of the individual potential fungal players, which is less sensitive to occasional changes in carbon or nitrogen supply.

To the best of our knowledge, we reveal for the first-time isotope signatures of hyphae of arbuscular mycorrhizal fungi in mycoheterotrophic plants and, together with Zahn et al. (2023), of fungal pelotons present in chlorophyllous orchids in relation to the plant tissues from their roots. Arbuscular mycorrhizal hyphae have isotope signatures that allow a significant distinction in ε¹³C abundance in relation to reference plants. Subsequently, hyphae resemble the ε¹³C of the mycoheterotrophic plants, suggesting that these mycoheterotrophs gain carbon from the detected fungi. For the orchid-associated fungi, hyphae are only slightly enriched in ¹³C in relation to both reference and orchid plants, remaining unclear whether these orchids gain C from the associated fungi based on these results. However, the significant enrichment in ¹⁵N of hyphae or orchid leaves indirectly indicates a partial mycoheterotrophic matter gain by these orchids.

Our study appeals to a careful interpretation when integrating root-associated fungal diversity and isotope natural abundances considering their inherent ecological significance as each method contains fundamentally different categories of information. Still, the combination of both approaches is greatly valuable and contributes to understand complex patterns in plant–fungal interactions, for example considering spatial and temporal fungal colonization in roots, and both advantages and caveats of each technique should be considered in the subsequent interpretation of ecological patterns. Finally, including the isotopic signatures of root-associated fungi in the context of mycorrhizal symbiosis contributes to a direct observation of fungal participation to organic matter gain of the plant.

None declared.

SIFG and GG designed the research and collected the orchid material. PG and SK guided the fungal hyphal extraction by SIFG. CH and KS collected the arbuscular mycorrhizal plant material. GG supervised the isotope abundance analyses. SIFG performed the molecular analysis, analysed the data, and together with GG wrote the manuscript. All authors commented and approved the final version of the manuscript.