Analyses of SNPs and genes clustered wood-degrading magic mushrooms from the northern hemisphere among Australian populations
We included 86 haplotypes from at least 28 separate mushrooms in Australia, with some haplotypes collected as populations and not associated with a single pileus. Our sampling covered 12 sites in eastern Australia (Fig. 2A) and included two reference genomes ofP. azurescens and one of P. cyanescens from the USA. kSNP called 1,580,296 SNPs, and we tested for population structure based on ancestry of 6,757 LD-corrected SNPs, which was a subset that excluded all sites that contained indels or missing data (Figs 2B, 2C). Mixed ancestry within sites and among known siblings appeared in DAPC analyses beyond K=7 (Fig. 2C). DAPC analyses showed populations were admixed in geographic locations. Psilocybe azurescens and P. cyanescens clustered with P. subaeruginosa in 2D plots (Fig 2B) and had recent shared ancestry with Australian populations (Fig. 2C).
We visualised relationships among genomes of P. subaeruginosawith SplitsTree neighbour networks based on 1,555,848 LD-corrected SNPs, including indels (Fig. 2D), and 194 aligned protein coding genes (76,076 amino acids, Fig. S1) identified by OrthoFinder. Relationships recovered by SNPs and genes were congruent. Psilocybe azurescens andP. cyanescens clustered among Australian populations.
Most individuals in populations from Bunya, Clifton Hill, Ellendale, kunanyi, Ravensbourne, and Shelley were sampled as siblings that could be linked to the same pileus (Fig. 2D). We used the AJK statistic (Yang et al., 2010) to investigate the relatedness of haplotypes within populations (Fig. S2). The observed relatedness suggests that haplotypes in clusters observed in Figure 2D are as related as known siblings even if they were not from the same pileus. These close relationships are supported by likelihoods of the AJK statistic ≥0.91, which is the lowest likelihood for known siblings sampled from Ellendale, Tasmania.
Groups defined by DAPC and supported by network analyses of SNPs and genes show that P. subaeruginosa is structured by geography in Australia. Relationships among groups reflect geographic boundaries, for example, samples east of the Great Dividing Range (which divides the eastern coast of Australia), namely Khancoban, Shelley, Clifton Hill, and Geelong, differed from populations west of the range in South Australia, Tasmania, and central Victoria. Full sib haplotypes sampled from one spore print from Clifton Hill (Fig. 2D e) were completely intermixed with sibling haplotypes (based on genetic distance and the AJK statistic) of other fruiting bodies from planted garden beds in both Clifton Hill and Geelong. A possible explanation for this is that a parental, dikaryotic genotype has spread long distances via woodchips and mulch in Victoria, facilitated by the perennial nature of P. subaeruginosa mycelium. Geographic areas with mixed ancestry based on DAPC (Fig. 2C) and genetic distance (Fig. 2D), namely Clifton Hill and Shelley, indicate that pilei were sampled from fruiting mycelia of different genotypes at the same location.