Premises

  1. (1)

    Definition of symbiosis. Symbiosis is a commonly used term in biology, but traditionally has two distinct meanings1. In what follows, we use it in the sense of de Bary2 to refer to the close and long-term coexistence of two different organisms. Symbiosis in this sense need not involve any beneficial or harmful relationship, merely close co-existence.

  2. (2)

    Definition of lichen. Lücking et al.3 provided a chronologically ordered list of lichen definitions. None of these is entirely satisfactory for our purposes, so here we define a lichen as follows. A lichen is an association of a fungus (mycobiont) and an alga or cyanobacterium (photobiont) with the following characteristics: (i) The mycobiont is nutritionally dependent on its photobiont. (ii) The mycobiont is not obviously harmful to its photobiont. (iii) The photobiont occurs within the mycobiont thallus. (iv) Mycobionts and photobionts usually cannot persist over a long period outside the symbiosis.

Introduction

Mutualistic relationship between photoautotrophs and fungi arose many times in evolution, had paramount importance in the development of terrestrial life and still remains essential in the present-day ecosystems4. A flagship of these relationships is lichen symbiosis, a highly elaborated cooperation between fungi and green algae and/or cyanobacteria where the fungal partner is nutritionally dependent on its photoautotroph5. Lichen symbiosis has multiple independent origins6 and its complexity and the stage of lichenisation differs considerably in various examples7. In some cases, a single fungal species may be either lichenised or saprophytic depending on conditions8,9, and the same is truth for algae10.

Numerous fungi are apparently associated with, and nutritionally dependent on algae, but their thallus is inconspicuous, not stratified into the typical lichen thallus6. A list of such fungi, so called “semilichens”, has been recently provided by Vondrák et al.11. Apart from semilichens, other algal-fungal symbioses that do not fully meet the definition of a lichen exist12. A remarkable but overlooked one, is linked to corticioid fungi, traditionally defined as basidiomycetes with fruiting bodies (basidiomata) appressed to the substrate with superficial non-poroid hymenium (extended definition in13). These flat basidiomata (“crusts” in further text) usually cover wood or bark.

Albertini & Schweinitz14 described one of the wood-dwelling corticioid fungi as Hydnum bicolor, currently named Resinicium bicolor. The epithet bicolor reflected the contrast between the white surface of fruiting bodies and the rusty brown tips of hymenial spines (the coloration frequently observed in older basidiomata). This epithet additionally matches an even more distinctive colour contrast—the white (or partly translucent) fungal crust is regularly green beneath. The green tinge is caused by an algal layer formed directly below the white fungal coat.

A detailed investigation of various corticioid fungi in European woodlands revealed living algal cells thriving below or inside the crusts of several unrelated fungal species. We propose to name these alliances “alcobioses” (singular “alcobiosis”), such as alga and corticioid fungus in symbiosis. Whereas some alcobioses form unstable associations where the algal cells are few in scattered colonies, others apparently have a tight relationship where algae form a lichen-like algal layer15. Only a few algal taxa have previously been reported in an association with corticioid fungi, mainly unicellular members of the green-algal class Trebouxiophyceae such as Coccomyxa glaronensis15 and undetermined species of Coccomyxa and Elliptochloris16,17. These algae are ubiquitous in terrestrial habitats and exhibit a general tendency to enter lichen-like symbioses18. However, their symbiotic and free-living members are morphologically uniform, and separation of the cryptic lineages requires molecular data19,20 not available for alcobioses yet.

The similarity of alcobioses and crustose lichens is remarkable, but the former have received little and only superficial attention. The question of whether alcobioses have a nutritional character, as in lichens21, has not been addressed. Here we provide the most comprehensive morphological and taxonomical assessment so far concerning both symbiotic partners. We also demonstrate that algal cells in these consortia are alive and metabolically and photosynthetically active even when fully embedded in the fungal crust. And we also studied carbon transfer from algal polyols (ribitol and sorbitol in our cases) into fungal mannitol that would confirm the nutritional relationship of the fungus to the algae.

Results

Alcobioses are stratified systems with an internal algal layer

Algal cells were observed enclosed either in the lower part of crustose basidiomata (subiculum) or in the substratum below the crusts, mostly rotten wood, however they never covered the crust surface. The density of algal cells varied considerably among infraspecific individuals and among species from an entire absence to a distinct thick continuous layer. The thickness of the algal layer varied, but frequently exceeded 100 µm. Whereas the algal layer was formed only occasionally in some species (Exidiopsis calcea and Tubulicrinis subulatus), it was found regularly in Lyomyces sambuci (Fig. 1), Resinicium bicolor (Fig. 2), Skvortzovia furfuracea (Fig. 3) and some Xylodon spp. All these fungi, however, were also recorded living separately, without an internal algal layer. In all cases, the colonies of algal cells were enclosed in fungal tissue (although they are also found in the substrate below the crust Fig. 3E). Nevertheless, a truly close contact where algae are encircled by fungal hyphae was mostly not observed. The only exception was Lyomyces sambuci which occasionally formed spherical goniocyst-like structures (20–40 µm in diameter) where a group of algal cells was tightly enclosed within the hyphal network (Fig. 1F).

Figure 1
figure 1

Association of Lyomyces sambuci and Desmococcus olivaceus (GPS: 48.9409975N, 14.5175219E; voucher: PRA-JV25262). (A) bark of Sambucus nigra covered by a free-living Desmococcus algal crust which is largely overgrown by Lyomyces; (B) vertical section of the Lyomyces crust with a distinct algal layer; (C) vertical section with the red chlorophyll autofluorescence; (D) algal colonies incorporated in a loose hyphal tissue, below the cover of compact fungal tissue; (E) Desmococcus in the algal layer; (F) Desmococcus and Lyomyces form lichen-like goniocysts. Scales: (B, C), 100 µm; (D, E, F), 20 µm.

Full size image
Figure 2
figure 2

Association of Resinicium bicolor and Coccomyxa (GPS: 48.6679583N, 14.7053083E; voucher: PRA-JV25257). (A) typical habitat–vertical surface of rotten spruce trunks; (B) R. bicolor crust; (C) vertical section with a distinct algal layer below the fungal coat; (D) the red chlorophyll autofluorescence indicates locations of Coccomyxa cells in the vertical section. Scales: (B) 5 mm; (C, D) 50 µm.

Full size image
Figure 3
figure 3

Association of Skvortzovia furfuracea and Coccomyxa (GPS: 48.6679583N, 14.7053083E; voucher: PRA-JV25255). (A) typical habitat – shaded surface of rotten spruce trunks; (B) S. furfuracea crust grazed by snails; (C, E) vertical sections of S. furfuracea crust. Fungal tissues coloured by lactoglycerol cotton blue. A distinct algal layer is visible below a dark blue fungal coat; (D) the red chlorophyll autofluorescence indicates locations of Coccomyxa cells in the vertical section; (F) Coccomyxa loosely integrated in the fungal tissue. Scales: (B) 1 cm; (C, D) 50 µm; (E) 20 µm; (F) 10 µm.

Full size image

Diverse fungal partners are associated with several ecologically distinct algae

The fungal partners appear to be more diverse in alcobioses than the associated algae. We found nine fungal species dispersed across the phylogeny of Agaricomycetes (Table S1, Fig. S3) and only three algal partners from the class Trebouxiophycae (Table S1, Fig. 4, Figures S4, S5, S6). The vast majority of fungi involved (i.e. seven species) entered the symbiosis with a single Coccomyxa species (Figures S4, S6A,B,C,D). All these fungi have similar ecology, occurring in temperate forests on decaying wood, especially on fallen spruce trunks, in shaded conditions (Figures 2 and 3; Table S1). The associated Coccomyxa sp. has a rbcL sequence almost identical (> 99% identity) to a lichen photobiont of Sticta22 and to a symbiont in an allegedly lichenised Schizoxylon albescens9. Both symbioses have quite distinct ecology: the former is a tropical lichen from Cuba and the latter is known from aspen bark in Northern Europe. In addition, the same algal genotype labelled Coccomyxa sp. “cort06” in terms of the rbcL sequence (~99.8% identity) was found to colonize the bark of trees with no report of association with fungi23.

Figure 4
figure 4

Linkage between algal and fungal partners in alcobioses. Algae are arranged in rbcL phylogenetic trees of Trebouxiophyceae; only parts with Stichococcus s.lat., Desmococcus and Coccomyxa depicted. Fungi are arranged in the ITS tree of selected Agaricomycetes. Symbionts in alcobioses are in grey rectangles. Detailed trees are available on Figures S3, S4, S5.

Full size image

The single fungus, Lyomyces sambuci, was regularly associated with another green algal species Desmococcus olivaceus, as determined both by morphological identity (Fig. S6E,F,G,H) and high rbcL sequence similarity (~99.5%; Fig. S5) to a typical strain of this species—SAG 1.9424. This alcobiosis differs in ecology from the associations with Coccomyxa. It is more tolerant of drying, occurs in lighter sites and has a strong affinity to bark and wood of living or dying Sambucus shrubs. The alliance Lyomyces-Desmococcus is apparently the most intimate among the observed alcobioses as it forms goniocyst-like structures and the algal carbon is undoubtedly transferred to the fungus (see below). However, again, the same species of Desmococcus also occurs free-living (Fig. 1A,24) and its symbiosis with Lyomyces is clearly facultative.

Finally, a green alga matching Tritostichococcus coniocybes, both in morphology (Fig. S6H,I) and the rbcL sequence (97–99% identity; Fig. S5), was detected in alcobiosis with a single fungus, Kneiffiella abieticola. It was observed on soft rotten wood of spruce snag in a microsite sheltered from rain. Tritostichococcus coniocybes is a lichen photobiont detected in Chaenothecopsis spp.24, Arthonia thoriana and Chaenotheca spp. (Fig. 4), and also was observed free-living (our data). It always occurred in rain-sheltered microhabitats.

Algae thrive in the association

Absolute fluorescence intensity (e.g. Fo) is very variable and demonstrates heterogeneity in algal chlorophyll abundance (Fig. 5A,D,G). The maximal quantum yield of photosystem II (Fv/Fm) is very homogeneous (0.55 to 0.75) for well hydrated alcobioses studied (Fig. 5, Fig. S7). Moreover, Fv/Fm for alcobioses and uncovered algal layer does not differ substantially (Fig. 5C,F,I). In contrast, dry systems have Fv/Fm close to zero and recover quickly after rehydration (Figures S7, S8). Minimal fluorescence (Fo) was increased to some extent when fungal crusts were removed by razor blade from algal layer. It may demonstrate a shielding effect of the fungal partner on algae, particularly in Lyomyces-Desmococcus system (Fig. 5J,K,L,M,N,O).

Figure 5
figure 5

Chlorophyll fluorescence imaging of fungal-algal associations. Minimal fluorescence (Fo), visual frame and maximal quantum yield of photosyntem II (Fv/Fm) for Resinicium bicolor-Coccomyxa (A, B, C, J, K), Skvortzovia furfuracea-Coccomyxa (D, E, F, L, M) and Lyomyces sambuci-Desmococcus (G, H, I, N, O). Each sample contains both algal patches and alcobiosis where fungal crust completely covers the algae. Bottom pairs of frames demonstrate shielding effect of fungal crusts. Upper frames are intact alcobioses (J, L, N), whereas fungal part of the system was carefully removed by razor blade in red frames of the bottoms (K, M, O).

Full size image

CO2 exchange of alcobioses was also very variable but easily detectable, and it confirms the viability and physiological activity of the partners. Typical light response curves of CO2 assimilation for five alcobioses, free-living terrestrial alga and foliose lichen are in Fig. S9. The following features were common for all systems: (1) Respiration of just rehydrated alcobioses was much higher (typically five to tenfold) than of those hydrated for a long time, equilibrating to steady-state in hours to days. (2) All alcobioses respond to light, and photosynthesis exceeded overall respiration in some cases (particularly under elevated [CO2]). (3) Photosynthesis is saturated under unusually dim light (typically 50 to 100 µmol photons m−2 s−1) in systems with Coccomyxa algae, but may be still increasing in the others, under full sun intensity (ca 1800 µmol m−2 s−1) and elevated CO2, Fig. S5). Moreover, lower temperatures are beneficial for the net carbon gain (Fig. S10).

Algal carbon is absorbed by fungi in some alcobioses

Pilot (HPLC–MS, GC–MS) experiments did not detect any significant carbon transfer from algal polyols (sorbitol and ribitol) to fungal substances (mannitol and ergosterol) in Resinicium bicolor-Coccomyxa, Skvortzovia furfuracea-Coccomyxa and Xylodon asper-Coccomyxa. In contrast, the control lichens, Hypogymnia physodes and Multiclavula mucida, expressed a clear pattern of 13C transfer from algal ribitol to fungal mannitol after two hours of assimilation in the 13C-enriched air. Intermediate results were obtained for Lyomyces sambuci-Desmococcus where mannitol expresses some degree of 13C enrichment, but barely significant in our set-up (data not shown).

Subsequent detailed studies using isotope ratio mass spectrometry (IRMS), much more sensitive for isotope abundances, were performed (Method S1). These experiments delivered two strikingly different results: (1) The Skvortzovia furfuracea-Coccomyxa system did not display any algal-fungal carbon transport. The principal polyols were algal ribitol and fungal mannitol (Fig. 6A). TMS-ribitol was highly 13C enriched (3.90 ± 0.26 At%, t = 24.4, P < 0.001, N = 5) after 18 h of labelling whereas the fungal TMS-mannitol remained very close to the natural 13C abundance (1.07 ± 0.03 At%, t =  − 0.8, P = 0.45, N = 5; Fig. 6D). (2) The algal-fungal carbon transfer was confirmed in the Lyomyces sambuci-Desmococcus alliance, where sorbitol is the principal algal polyol (Fig. 6C). We recorded a substantial 13C signal in TMS-sorbitol (2.15 ± 0.07 At%, t = 28.8, P = 0.001; N = 4) and a slightly lower but still very significant signal in TMS-mannitol (1.48 ± 0.10 At%, t = 8.2, P = 0.004, N = 4; Fig. 6F).

Figure 6
figure 6

Abundance and 13C enrichment in trimethylsilyls of principal algal and fungal polyols after 18 h assimilation of alcobioses in 13CO2 atmosphere. Algal polyols: ribitol (R) and sorbitol (S); fungal mannitol (M). Median (central point), 25% and 75% quantiles (box) and min–max (whiskers) are shown. Dotted line represents natural 13C abundance (1.07 At %) and t-test was performed against it; NS: not significant (P > 0.05), *: 0.05 < P > 0.01, **: 0.01 < P > 0.001, ***: P < 0.001.

Full size image

Two specimens of Botryobasidium-Coccomyxa, measured in parallel, having low biomass and, thus, low polyol abundances (Fig. 6B), delivered very convincing negative result. Its TMS-ribitol 13C enrichment was high and consistent (3.70 ± 0.14 At%, t = 25.9, P = 0.024; N = 2) but mannitol invariant from natural abundance (1.11 ± 0.04 At%, t = 1.4, P = 0.39, N = 2; Fig. 6E), very similar to Skvortzovia furfuracea-Coccomyxa system. See also chromatograms (Figures S11, S12).

Snails rejuvenate alcobioses and produce isidia-like diaspores

Literature says little about the persistence of the corticioid fungal crusts involved in this study. According to our field observations, crusts of Lyomyces sambuci, Resinicium bicolor and Skvortzovia furfuracea can persist over several years. Persistence of alcobioses is apparently assisted by snail grazing. Large areas of fungal crusts including algae are frequently removed by snails and the algal layer is uncovered (Fig. 7A). Grazed spots are quickly overgrown by rejuvenated fungal hyphae (Fig. 7B). Therefore, fungal crusts, especially of Skvortzovia furfuracea, typically form a continuous mosaic of younger and older (i.e. thinner and thicker) patches.

Figure 7
figure 7

Snail-grazed Skvortzovia furfuracea. (A) a green area of exposed algal layer formed by snail grazing; (B) regenerated mycelium overgrowing grazed areas. Algal cells are indicated by the red chlorophyll autofluorescence. (C) fresh snail excrement; (D) excrements are quickly overgrown by regenerated mycelium; (E) the red chlorophyll autofluorescence indicates high density of Coccomyxa cells in an excrement; (F) a mixture of Coccomyxa and remnants of fungal tissue in an excrement. Scales: (A) 2 mm; (B) 0 µm; (C, D) 1 mm; (E, F) 20 µm.

Full size image

Snail excrements form distinct structures accompanying alcobioses. They have a granular or caterpillar-like shape (Fig. 7C) and, when lying on grazed fungus, they are quickly overgrown by young hyphae and incorporated into the fungal crust (Fig. 7D). Fresh excrements are green, densely filled by living algal cells (Fig. 7E) in a mixture with remnants of fungal hyphae (Fig. 7F). The algal-fungal content makes excrements an analogous structure to vegetative propagules (e.g. isidia) of lichens and may serve for vegetative reproduction when being placed/translocated outside the current fungal crust. The same function was described at mite excrements that contained viable fungal and algal cells of the lichen Xanthoria parietina25.

Discussion

Corticioid fungi associated with algae

Corticioid fungi are generally considered saprophytic, rarely parasitic and mycorrhizal; the frequent association with algae has largely been ignored in most monographs26. However, Parmasto27, already in 1967, observed an alcobiosis in the newly described fungus Phlebia lichenoides, as he refers to an internal algal layer containing Chlorococcus-like cells. Hjortstam et al.28 considered Phlebia lichenoides as a synonym of P. subcretacea (currently Cabalodontia subcretacea) and dismissed its association with algae.

Poelt & Jülich15 provided the most comprehensive study so far on alcobiosis revealed in Resinicium bicolor, and refer to the presence of an algal layer formed, allegedly, of Coccomyxa glaronensis. (This algal species was described as a symbiont in the lichen Solorina saccata29). Poelt & Jülich15 observed algae in all surveyed specimens and algal colonies were either restricted below fungal crusts or only slightly expanded out of the fungal spots and formed a surrounding green rim (which corresponds with our observations; Fig. 5A,B). The authors classified the algal-fungal contact as intimate, but without obvious appressoria or haustoria. Subsequent short notes in literature13,30,31,32 have only repeated the findings by Poelt & Jülich. Whereas Oberwinkler speculated in 1970 that R. bicolor represents a basidiolichen30, he omitted this fungus from his later overview of basidiolichens33.

Two related recent conference contributions refer to associations of corticioid Hyphodontia s.lat. (namely Lyomyces crustosus, Hyphodontia pallidula and Xylodon brevisetus) with algae Coccomyxa and Elliptochloris, but without molecular sequence data16,17. It is not clear from the published meeting abstracts, if both algal genera were observed co-occurring in an association with a single fungal crust, or if each algal-fungal system involved only a single algal partner. We have only observed the latter case. The anatomical observations reported by Voytsekhovich et al.17 allegedly revealed appressoria and haustoria in hyphae attached to algal cells and, on this basis, the authors consider these fungi to be optionally lichenised. Physiological relationships between the corticioid fungi and their algae were not studied. Our data confirmed that some species of Hyphodontia s.lat. are frequently involved in alcobioses, but we do not consider them lichens (see below). Whereas Coccomyxa is undoubtedly frequent in alcobioses, we did not detect Elliptochloris in specimens collected in the current study.

Associations of corticioid fungi and algae sometimes have a typically parasitic character. It was demonstrated for Athelia epiphyla15 where fungal hyphae form haustoria penetrating algal cells. This fungus causes bleaching of corticolous algae, i.e. forms characteristic pale-grey rounded spots on otherwise green tree bark. Although A. epiphyla is apparently a parasite, Jülich34 and Oberwinkler33 refer to symbiotic relationships between epiphytic algae and some species of Athelia and Athelopsis. Some of these cases may be close to our concept of alcobiosis.

Parallel research has been conducted on associations of polypore fungi and their epiphytic algae. The surface of polypore fruiting bodies is a suitable substrate for numerous algal (and cyanobacterial) species35,36. However, these associations do not show any specificity–a single polypore is often covered by a community of several algal species with broader niches (not restricted to polypores). Carbon transfer from epiphytic algae to polypore fungal tissues was repeatedly reported using 14C tracer37,38, but we are skeptical of that result as the algae usually grow on/in dead polypore tissues. Thus, this association seems to be very different from alcobioses with algae embedded within the viable fungal tissue.

The taxonomic composition of algae found in alcobioses is in general not surprising. Unicellular trebouxiophyte algae such as Coccomyxa and Stichococcus sensu lato are ubiquitous in terrestrial habitats and have been long known to create various types of fungal-algal consortia39. In our study we employed molecular barcoding of the algal partners in alcobioses for the first time, which allowed us to place all three detected taxa into narrowly defined and strongly supported monophyletic clades of Coccomyxa sp., Desmococcus olivaceus, and Tritostichococcus coniocybes. Each of these clades probably corresponds to a single algal species, judging from their uniform morphology and high rbcL gene sequence identity (Fig. 4, Figures S3, S4, S5). The Coccomyxa sp. found in our samples evidently represents a species new to science, but its transfer to pure culture, deposition of a holotype, and formal description was beyond the scope of the current study. As well as living in alcobioses, we know from previously published sequences that each of these three genospecies can live as a free-living terrestrial alga23,24. In the case of D. olivaceus, our results are the first reliable observation of this species in symbiotic association with fungi, but S. coniocybes and the particular clade of Coccomyxa have previously been reported as genuine lichen photobionts22,24.

Ecophysiological implications

We confirmed the viability of algae in alcobioses via chlorophyll a fluorescence40 and gas exchange measurement. The maximal quantum yield of photosystem II (Fv/Fm) is a well-accepted measure of vitality and impact of stress. In the case of alcobioses, Fv/Fm confirms that the algae are not stressed under the fungal layer, having Fv/Fm comparable to values for adjacent free-living algae (Fig. 5C,F,I). Photochemistry recovered within minutes to tens of minutes to values > 0.4 after dry alcobioses were rehydrated (Figures S7, S8), being considered “physiologically active”, in accordance with studies on terrestrial algae41.

The minimal fluorescence (Fo), in turn, shows a low shielding effect of fungal crusts to incident light, particularly for shade-adapted alcobioses formed by Coccomyxa and fungi with rather thin basidiomata (Fig. 5J,K,L,M). A substantial shielding effect was detected at the thicker and more light-scattering crust of Lyomyces sambuci, growing in more-light exposed sites, which may be photoprotective for Desmococcus algae in this association (Fig. 5 N,O).

The CO2 exchange, reflecting the real photosynthetic performance and production process, has not been studied in alcobioses yet, but reference data are available for lichens42,43,44. Here we provide gas-exchange data for four alcobiosis systemes in the context of lichen symbiosis (represented by Parmelia sulcata) and a free-living algal crust (Trentepohlia aurea). Figures S9 and S10 demonstrate relationships of the CO2 exchange to an incident light intensity, gaseous CO2 concentration and ambient temperature. Respiration of the fungal partner plus co-occurring microbiomes was usually higher in shade-adapted Coccomyxa-based alcobioses (Fig. S9). Thus, net carbon balance of the system was mostly negative and algal photosynthesis could not serve as a principal source of carbon for the fungus. Saturating light intensity was also unusually low (50 to 100 µmol photons m−2 s−1) suggesting a strong shade adaptation of the algae involved. In contrast, free-living Trentepohlia aurea, the lichen Hypogymnia physodes and the alcobiosis Lyomyces sambuci-Desmococcus occurring in more light-exposed conditions, had higher maximal CO2 assimilation and higher light saturation intensity and the carbon gain of these systems was frequently positive. CO2 assimilation rate rising even close to full sun intensity under elevated CO2 suggests high diffusional limitation of the photobionts and their high photosynthetic capacity as well (Fig. S9). In addition, the overall carbon balance of those systems is strongly temperature dependent. Lower temperatures promote carbon gain of alcobioses, as demonstrated on Xylodon system (Fig. S10). This can be explained by higher temperature dependency of dark respiration (Rdark) than gross photosynthetic capacity (Agross). Comparable magnitudes of Rdark and Agross will lead to substantial effect of temperature on carbon gain45.

We observed a significant alga-to-fungus 13C transfer in only one of the systems studied. We selectively measured algal (ribitol, sorbitol) and fungal (mannitol) compounds. Polyols are not suitable for gas chromatography, therefore their trimethylsilyls (TMS-) were measured46. That is, the five-carbon ribitol was converted to the twenty-carbon TMS-ribitol (C20H52O5Si5). Similarly, six-carbon mannitol and sorbitol led to the formation of 24-carbon TMS-derivatives (C24H62O6Si6). Thus, the 13C enrichment (indicating carbon transfer) is four-times more pronounced in mother molecules than in TMS derivatives measured here and thus the negative result for Skvortzovia furfuracea-Coccomyxa is very convincing (Figs. 7 and S11). The isotopic precision of IRMS is better than 0.01 At% of 13C for isotope ratios close to natural (< 5 At% of 13C). Mannitol 13C is clearly unchanged from its natural value in this system, whereas TMS-ribitol is highly enriched up to 4.6 At% 13C (by 3.5 At% compared to natural). This means that the parent ribitol should have up to 15.1 At% 13C (4 × 3.5 + 1.1). In contrast, the Lyomyces-Desmococcus TMS-mannitol was significantly enriched by approximately 0.4 At% (Figs. 7 and S12). Thus, mannitol should be enriched up to 2.7 At% (1.1 + 4 × 0.4 At%). Despite the clear statistical significance, more work is needed to decipher the timing, environmental dependencies and, thus, ecological significance of those carbon transfers.

Symbiosis on the threshold of lichenisation

The lichen is defined above in the Premise 2 as a symbiosis of alga or cyanobacterium (photobiont) and fungus (mycobiont) with following specifics: (1) The mycobiont is nutritionally dependent on its photobiont47. (2) The mycobiont is not obviously harmful to its photobiont48. (3) The photobiont occurs within the mycobiont thallus49. (4) Mycobionts and photobionts usually cannot persist over a long period outside the symbiosis50,51.

Alcobioses do comply with points (2) and (3). They have an internal lichen-like algal layer (Figs. 1, 2, 3) and the algae thrive in the symbiosis (Fig. 5, Figures S3, S4, S5, S6). Only Lyomyces-Desmococcus partly complies with point (1); we confirmed carbon transfer from alga to fungus (Fig. 6, Figures S11, S12). We suggest however that Lyomyces is not fully dependent on algal assimilates, because it has been observed occasionally without the algal symbiont. Consequently, Lyomyces-Desmococcus does not meet point (4), because both symbionts may live apart. Surprisingly, Desmococcus olivaceus is absent from the list of lichen photobionts39 and is mostly reported free-living, forming extensive epiphytic or epilithic algal crusts52. Another member of the genus, Desmococcus vulgaris, was found overgrowing fruiting bodies of the polypore Fomes fomentarius53. Lyomyces sambuci, like other Hyphodontia s. lat., is believed to cause white rot, i.e. is saprophytic and, remarkably, the frequently present algal layer in this common fungus was not mentioned or illustrated in the monographs26,28. Point (4) is partly met in some alcobioses with the Coccomyxa species known from lichens9,22. At our sampling sites, the alga appears to occur only inside and below the fungal crusts (Fig. 5A,B). Contrary to our observations, an almost identical genotype of Coccomyxa was found free-living on the bark of live pine and oak trees in the study of corticolous algae by Kulichová et al.23. Therefore, the lifestyle of this alga seems to be diverse, and investigations on the population level are needed to elucidate the potential specific life strategies of individual strains. Fungi associated with Coccomyxa were observed to live without algae, but the alcobiosis is almost omnipresent in Resinicium bicolor and Skvortzovia furfuracea. The nature of this symbiosis remains enigmatic as it probably has no nutritional character.

In conclusion, the photosynthetic potential of algal partners is clearly substantial, but their direct nutritional importance for fungi (and whole alcobioses) is still obscure, even in Lyomyces-Desmococcus, and needs more study. Simultaneously, the insignificant carbon exchange observed in most systems implies that there must be other ecological advantages keeping the partners in a stabile association (e.g. exchange of bioactive substances under stress conditions such as drought).

Materials & methods

In the period 2016–2021, we recorded 58 specimens with alcobiosis (Table S1). Corticioid fungi were identified to species from their morphology and 27 specimens were barcoded by ITS nrDNA sequences. Algae were determined to genus according to their morphology and 20 specimens were barcoded by sequencing the ribulose bisphosphate carboxylase large subunit gene (rbcL). The NCBI accession numbers of the sequences obtained are provided in Table S1. Field collections from 2019–2021 were employed in morphological and anatomical observations (fluorescent microscope Olympus BX 61 in bright field and fluorescent mode) and in physiological measurements that were done within one week after collection. Before that, samples were hydrated if necessary, put in Petri-dishes and accommodated in LED illuminated cultivating room under 20 °C, irradiation about 100 µmol m−2 s−1 and photoperiod of 12 h for at least two days.

DNA barcoding & phylogenetic analysis

Fresh specimens were used for DNA extraction of both fungal and algal partners. DNA was extracted with a cetyltrimethylammonium bromide (CTAB)-based protocol54. ITS nrDNA locus was found to be a useful fungal barcode sequence, easily amplifiable, and moreover with a sufficient number of references in the NCBI database. In the case of algae, we sequenced nrDNA ITS and 18S regions and the rbcL gene. The latter was found as best amplified and therefore selected for barcoding and phylogenetic analysis. Polymerase chain reactions were performed in a reaction mixture containing master mix consisting of 2.5 mmol/L MgCl2, 0.2 mmol/L of each dNTP, 0.3 μmol/L of each primer, 0.5 U Taq polymerase in the manufacturer’s reaction buffer (Top-Bio, Praha, Czech Republic), and milli-Q water to make up a final volume of 10 μL. The primers used for PCR and the cycling conditions are summarized in Table S2. Successful amplifications were sent for Sanger sequencing (GATC Biotech, Konstanz, Germany). Sequences were edited using BioEdit v.7.0.9.055 and Geneious Prime 2022.0 (https://www.geneious.com).

Sequences of our specimens were supplemented by relevant sequences from GenBank—NCBI database. Sequences were aligned by MAFFT v.756; available online at http://mafft.cbrc.jp/alignment/server/) using the Q-INS-i algorithm and adjusted manually. The best-fit model of sequence evolution was selected using the Akaike information criterion calculated in jModelTest v.0.1.157. Relationships were assessed using Bayesian inference as implemented in MrBayes v.3.1.258. Two runs starting with a random tree and employing four simultaneous chains each (one hot, three cold) were executed. The temperature of a hot chain was set empirically to 0.1, and every 100th tree was saved. The analysis was considered to be completed when the average standard deviation of split frequencies dropped below 0.01. The first 25% of trees were discarded as the burn-in phase, and the remaining trees were used for construction of a 50% majority consensus tree.

Gas exchange

CO2 exchange was measured using the portable photosynthetic system LI-6400XT (Li-Cor, Lincoln, NE, USA) connected to a custom-made peltier-conditioned gas exchange chamber described elsewhere59,60, see also Figure. S1. The chamber allows to accommodate samples up to 64 cm2 of ground area and 1 cm in thickness. It is of high sensitivity, as we have demonstrated in the past by successfully measuring the very small gas exchange of moss capsules61. Algae-containing basidiomata were attached to the razor-thinned native substrate to minimise microbial respiration. “White” PAR irradiation was supplied by a LI6400-18 RGB system source controlled by LI-6400XT. Temperature was maintained at 20 °C (except when measuring the response to varying temperature) and airflow at 250 µmol mol−1. The light response of CO2 assimilation was measured in continual-logging mode (each 5 s), at least 3 min in each light intensity. After a steady state was reached, data were collected, and light intensity changed to the next value. The temperature curve relied on Peltier-cooling of the cryptogamic chamber. At least 15 min was allowed to pass before data on gas exchange was taken, to allow equilibrium to be reached. Finally, background (empty chamber response) was measured and subtracted.

Chlorophyll a fluorescence imaging

2D fluorescence measurements were performed by the FluorCam FC800 instrument (PSI, Drásov, Czech Republic). Red LEDs (λ = 660 nm) were used for both, measuring light (10, 20 or 33 µs flashes, in average < 1 µmol m−2 s−1) and actinic light (continuous, photosynthesis driving, about 100 µmol m−2 s−1) photon sources. White LEDs (≈ 2500 µmol m−2 s−1) delivered saturation pulses of 1 s in duration. The camera has 720 × 560 px resolution, 12-bit data depth and is equipped with a zoom objective imaging frame down-to ca 10 × 7.5 cm (≈ 0.13 mm per pixel).

Metabolite transfer measurement

Stable carbon 13C was chosen to trace photobiont assimilated carbon. Substrate-attached sporocarps with active algal partner (Li-6400XT-proved, Figure. S1) were enabled to assimilate 13CO2 enriched air in a water-sealed 3L inverse Petri dish with internal fan (See Figure. S2). The labelling device is described in detail elsewhere61. [13CO2] was about 1000 µmol mol−1. Atmospheric CO2 was previously replaced by CO2 free synthetic air and then ca 3 mL of 13CO2 (> 99 atom %, Sigma-Aldrich, Luis., USA) was injected. Samples assimilated for two to 18 h under about 200 µmol m−2 s−1 of white LED light. Then fungal/algal crusts were scratched down by razor and killed in boiling methanol. Homogenised and filtered metOH extracts were used for analysis of polyols and ergosterol (see Methods S1 for further details).

We employed three approaches to trace 13C enriched metabolites. (1) HPLC–MS to separate algal polyols (ribitol and sorbitol) from fungal mannitol and to measure their 13C content. Although mannitol was detected in some algal groups62,63, it is not produced by algae in our systems, i.e., Coccomyxa and Desmococcus64. (2) GC-IRMS to separate trimethylsilyl derivatives of those polyols (TMS-ribitol, TMS-mannitol and TMS-sorbitol) and, again, to quantify their 13C enrichment. And (3) HPLC–MS to separate fungal specific ergosterol and to measure if it is 13C enriched. For more details see Method S1.