The effect of localities
Regarding locality, Benjamin Constant had the most differentiated community in all datasets (Fig. 4). The effect of localities was significant (p < 0.001) and had a higher effect than the soil layer factor in explaining the community composition in all datasets (18S – R2 = 0.10, COI – R2 = 0.12, and ITS – R2 = 0.11). In general, the pattern of highest mean, number of exclusive, and number of indicator OTUs by locality varied between markers (Table 3, Fig. 5). For 18S and ITS, Benjamin Constant had the highest mean number of OTUs (Table 3) and the highest number of exclusive OTUs (Fig. 5). Benjamin Constant also had the highest number of OTUs considered indicators of this locality for the 18S dataset (Table 3). Cuieras had the lowest number of exclusive OTUs for COI and 18S, but had the highest number of indicator OTUs for the COI datasets (Fig. 5). The majority of indicator OTUs at all localities were saprotrophs, followed by a high proportion of OTUs that could not be classified by their functional group (Table S2).