Mycorrhizal dependence in orchids ranges from partial mycoheterotrophy to obligate holomycoheterotrophy and from general to specialized associations. While orchid seed germination and plant distribution can be linked to the distribution and abundance of orchid mycorrhizal fungi (OMF) in soil, structural similarities between OMF communities in soil and those within the roots are not known. We tested the hypothesis that sites hosting orchid species will have high structural correspondence between the OMF communities in soil and in roots while soil at sites lacking the orchid populations will be devoid of OMF. By repeated sampling of soil and / or roots from eight sites across the entire natural range of Platanthera praeclara over the course of three years, we quantified and compared the structural composition of OMF recovered from orchid roots to soil OMF communities. Soil samples representing both orchid absence and orchid presence at each of the eight sites were collected in Iowa (1 site), Nebraska (1 site), Minnesota (3 sites), North Dakota (2 sites) in the United States, and in Manitoba, Canada (1 site). We used the nuclear ribosomal internal transcribed spacer2 (nrITS2) region for fungal community comparisons. While sanger sequencing was utilized for root OMF, high throughput illumina-based sequencing was used to quantify soil fungal communities.
ANOVA revealed differences in alpha diversity of OMF communities in soil samples collected across the eight sites (p = 0.00); however, the orchid presence/absence locations were indistinguishable (p = 0.09); the interaction between site and orchid presence/absence treatment also showed differences in OMF diversity (p = 0.004). The relative abundances of OMF families in soil samples showed Tulasnellaceae as the most abundant OMF family whereas Ceratobasidiaceae was the most abundant and dominant OMF family within the roots of this taxon at all sampling sites. PERMANOVA-based beta diversity comparisons with weighted and unweighted Bray-Curtis distances revealed differences in soil OMF abundance and richness among sites, orchid presence/absence treatments, and interaction of sites and orchid presence/absence treatment (p < 0.05 for all). Data suggest that the specificity of P. praeclara towards Ceratobasidiaceae OMF family may not be a consequence of the richness and abundance of OMF families in soil. Instead, the specificity might be a result of co-evolution of OMF species and its host plant. Further, the lower abundance of Ceratobasidiaceae family in soil might be the reason of rarity of P. praeclara.