Across all samples in the studied pine forest, 38% of the sequenced fungal community was assigned to class Archaeorhizomycetes in the short amplicon “ecological” dataset (Table S4). In the “phylogenetic” dataset on the other hand, the class represents 26% of the fungal reads (Table S2). Despite high relative abundance based on total fungal reads and intense cultivation efforts no isolates of Archaeorhizomycetes were successfully obtained. The relative abundance of class Archaeorhizomycetes was not significantly affected by soil horizon (Fig. S5, Table S5) or treatment (Table S5). However, number of Archaeorhizomycetes itASVs and PSHs were significantly affected by soil horizon (Table S6), with higher richness detected in B horizon samples compared to samples from both O and E horizon (Fig. S6). Across all samples, 68 PSHs of Archaeorhizomycetes were delimited, nine of which were supported by reads in the “phylogenetic” dataset, one of which included the reference sequence of A. finlayi (Fig. 1, Fig. S7). The most abundant PSHs were supported by long read data in the “phylogenetic” dataset (Fig. S4), together accounting for 78% of the Archaeorhizomycetes reads identified in the “ecological” dataset. Likely due to the lower sequencing depth of the “phylogenetic” dataset, taxa making up less than 1% of the sequenced fungal community were not consistently recovered in long read ASVs (Fig. S4).
The Archaeorhizomycetes community composition was structured by soil horizons (Fig. S8) and model testing showed that the relative abundance in the “ecological” dataset of nine PSHs supported by long reads was significantly affected by soil horizon (PERMANOVA: PSH vs Horizon df = 2, F = 4.6070, P < 0.0001) but not by treatment or plot (Table S6). The niche distribution also varied with plot*treatment interactions (df = 3, F = 2,0344, P > 0.005) (Table S6). This indicates that soil horizons may reflect niches explored differently by these fungi. While most of the nine PSHs were not significantly associated with a single horizon, we found that A. finlayi was tentatively associated with the O horizon (Table S7). For two sister pairs of phylogenetically well supported PSHs that were both abundant and frequently observed (Fig. S4, Table S8), we found significant differences in realized niche for PSHs within each pair (Fig. 1, Table S9). Based on their relative sequence read abundance in the fungal community across soil horizons, PSH_7 and PSH_8 had significantly different niche distributions (p < 0.05). Differential niche distribution of PSH_1 and PSH_2 was marginally significant (p < 0.1; Table S9). This provides further evidence that the phylogenetically distinct PSH_7 and PSH_8 are also ecologically separate species.
Global perspective on recognized taxa
The site-specific Archaeorhizomycetes diversity (Fig. 1) was analyzed in a global perspective by populating an alignment of ASVs from our “phylogenetic” dataset with publicly available environmental sequences that formed well supported clades with long reads Archaeorhizomycetes ASVs from the current study (Fig. 2, Supplementary datafile 6). A total of 78 global PSHs were delimited and supported using a bPTP model, 41 of which are represented by a single sequence (Fig. 2). However, some of the PSHs delimited in the local dataset (Fig. 1) were not stable with the addition of published environmental sequences. Most notably, PSH_2, PSH_5 and PSH_9 (identified as A. finlayi) each split in the global phylogenetic tree (Fig. 2). PSH_2 separates into two global PSHs, both containing sequences previously recovered from the studied field site, with PSH_2:2 being the more frequently observed of the two and including long sequences in UNITE SH1566367.08FU, while sequences in PSH 2:1 did not map to an existing UNITE SH (Supplementary datafile 6). Similarly, global PSH_5:1 contained previously published sequences from the same field site, as well as from Ireland and the US, while global PSH_5:2 contained only two ASVs from the current study (Supplementary datafile 6). Both global PSH_5:1 and PSH_5:2 cluster on a well-supported branch with four global PSHs that include previously published sequences (Fig. 2). Sequences in three of these PSHs map to the same UNITE SH, demonstrating that the boundaries of phylogenetically delimited PSHs and cluster-based SHs are not always the same. Further, ASVs and published sequences mapping to SH1556760.08FU, Archaeorhizomyces finlayi, split into two global PSHs (PSH_9:1 and PSH_9:2) in the current analysis (Fig. 2, Supplementary datafile 6). These two PSHs cluster on a well-supported branch with five single sequence PSHs mapping to four different UNITE SHs. Limited taxon sampling and low representation of intra species genetic variation of potential sister taxa as well as possible chimeric sequences within the clade may obscure PSH delimitation.
The included sequences represent only a fraction of the global Archaeorhizomycetes diversity, including sequences from 90 of the 181 UNITE SH currently identified as belonging to Archaeorhizomycetes (Kõljalg et al., 2013). With the exception of PSH_4 and PSH_2:1, long read ASVs mapped to UNITE SH meaning that the majority of our ASVs where highly similar to previously observed environmental sequences (Supplementary datafile 6). Archaeorhizomycetes borealis is distributed across the Eurasian boreal biome (Menkis et al., 2014) but was rare at our study site and was only detected as a short amplicon itASV.
In the global analysis, the sister taxa PSH_7 and PSH_8 cluster together with three single sequences (Fig. 2), two of which map to SH1566388.08FU (Supplementary datafile 6). Based on the globalfungi.com database (Vetrovsky et al., 2020) this SH restricted to North America, indicating that closely related taxa exist globally but that these are likely geographically separated from those captured in our dataset. While addition of publicly available environmental sequences weakened the bPTP ML support for these two PSHs, both remain intact. In the global tree, the delimitation of global PSH_7 and global PSH_8 both include sequences previously recovered from the studied field station as well as sequences collected throughout Europe (Supplementary datafile 7). These two PSHs are distinct based on both local (Fig. 1) and global (Fig. 2) phylogenetic analysis as well as ecological evidence demonstrating that they have significantly different realized niches across soil horizons at the study site (Fig. S9, Table S9). Overall, relative abundance of PSH_8 is higher at 8.9 ± 1,7% across all three horizons compared to PSH_7 at 3,5 ± 1,2%. Relative abundance of PSH_8 is highest in O horizon and decreases towards deeper soil layers while the relative abundance of PSH_7 is stable throughout the soil profile (Fig. S9a). When co-existing in a sample, relative abundance of PSH_7 is higher when the relative abundance of PSH_8 is low (Fig. S9b) indicating competitive avoidance between the two species similar to patterns of vertical separation previously observed for soil fungal sister species (Mujica et al., 2016).
Based on combined phylogenetic and ecological evidence we propose two novel species 1) Archaeorhizomyces secundus nom. sEq. for PSH_7 and 2) Archaeorhizomyces victor nom. sEq. for PSH_8 with names appended with nom. sEq. to indicate that the names are based on a sequence in the absence of acceptable type material.
Archaeorhizomyces secundus nom. sEq. Kluting, M. Ryberg & Rosling sp. nov.
Diagnosis: Separated from other species in the genus by ribosomal sequences possessing the following distinctive characters in ITS1: CCGAGTCGCCACAT, at position homologous to bases 103–116 in ASV_4; and in ITS2: CCATACCTTTTTGGTGTGT, at position homologous to bases 317–335 in ASV_4.
In DNA extracts from soil and from roots, often from ectomycorrhizal roots of Pinus sylvestris but also from roots of Calluna vulgaris. Found mostly in pine forest but also in other coniferous forest including spruce and heatland, in temperate, boreal and alpine climate. Found in both organic and mineral soil horizons but is less abundant in upper soil layers compared to Archaeorhizomyces victor.
The species is outnumbered by its sister species for colonization of organic soil.
ASV_4 based on sequence reads from Ivantjärnsheden field station, Jädraås, SWEDEN, Uppsala, October-2013, UDB0779127 in UNITE
Sweden, Norway, Finland, United Kingdom, Austria.
AB560514, DQ309209, FM992980, HM069470, JN032483, KX289979 (Alignment in Supplementary datafile 8).
Archaeorhizomyces victor nom. sEq. Kluting, M. Ryberg & Rosling sp. nov.
Diagnosis: Distinct from other species in the genus by ribosomal sequences possessing the following distinctive characters in ITS1: CGAATGGCTTTT at position homologous to bases 48–59 in ASV3, and ATGTGCTTTGGCGCCAAGT at position homologous to bases 93–111 in ASV_3; and in ITS2: TCATACCTTCTT at position homologous to 323–333 in ASV_3.
In DNA extracts from soil and from roots, often from ectomycorrhizal roots of Pinus sylvestris but also from Calluna vulgaris. Most frequently found in coniferous forests but also in deciduous forest, in temperate, boreal and alpine climate. Found in both organic and mineral soil horizons, outnumbers Archaeorhizomyces secundus in upper soil layers.
The species wins in competition with sister species for colonization of organic soil.
ASV_3 based on sequence reads from Ivantjärnsheden field station, Jädraås, SWEDEN, Uppsala, October-2013, UDB0779126 in UNITE.
Austria, Finland, Germany, the Netherlands, Sweden, United Kingdom.
AB560521, DQ309123, HM069408, HQ873359, JF300381, JN006467, JN006468, JN032485 (Alignment in Supplementary datafile 8).