Alder Soils vs Non-alder Soils
Differences between late succession soils derived from A.
tenuifolia rhizospheres and those derived from ‘non-alder’ areas were
much weaker than those between early and late succession alder
rhizospheres, at both the clade and sub-clade phylogenetic scales. Among
clades, the overall LR test was barely significant (P = 0.02,
pseudo-R2 = 0.02), with the largest difference
occurring in the proportional representation of ‘typical Alnus ’
sequences, which occurred in nearly 17% of non-alder clones compared to
6% of clones derived from A. tenuifolia rhizosphere soil (P =
0.007) (Table 1). Differences among sub-clades based on alder presence
were stronger (LR P < 0.0001, pseudo-R2 =
0.08) than among clades (Table 2). In addition to the aforementioned
difference in ‘typical Alnus ’ sequences, which was also
significant at the sub-clade scale (P = 0.01), the nodule-dwelling group
RF1_2_3_14 had significantly higher representation in alder than
non-alder soils (P = 0.02), although the absolute difference was only
four clones vs zero. Within the soil-only clades, several sub-clades
showed an affinity for either alder or non-alder soils. Among the former
were sub-clades A_4 (P = 0.002), B_1 (P = 0.001), and B_2 (P
< 0.0001). Among the latter, A_3 (P = 0.002) and B_4 (P
< 0.0001). Both alder and non-alder soils supported similar
OTU richness (Figure 2).