Richness and evenness between sampling strategies
Contrary to our predictions, richness differed only for root bacteria where homogenizing before subsampling resulted in slightly more SVs recovered than homogenizing after subsampling (Supporting Information, Table S2 and Fig. S1). Evenness did not differ for any microbial group. In addition, neither sampling strategy produced saturated species accumulation curves for any microbial community sampled, although AMF sampling approached saturation (Supporting Information, Fig. S2). This indicated inadequate sampling to characterize site richness, due to species turnover among plants. However, sequencing effort curves did saturate for all groups, indicating that sequencing depth was not a limiting factor in estimating the richness in individual samples (Figure 2).
The most abundant SVs, (those occurring in at least half of our samples and recovered by both sampling strategies), represented only a small proportion of the microbial communities recovered. This included 5.5% of all AMF SVs (Glomus, Clairoideoglomus ), while just 0.7 % of total bacterial SVs (Bacillus, unknown bacteria) and 0.9% non-AM fungal SVs (Nectriaceae, Plectosphaerella, Tetracladium ), fit these criteria. Foliar fungi were the most SV-rich group
(Figure 3), and only 0.8% of total SVs (Mycosphaerella ) occurred in at least half of all samples and were recovered by both sampling strategies.