ABSTRACT: Fungi underpin almost all terrestrial ecosystem functions, yet our understanding of their community ecology lags far behind that of other organisms. Here, red paddy soils in subtropical China were collected across a soil depth profile, comprising 0-to-10-cm- (0-10cm-), 10-20cm-, and 20-40cm-deep layers. Using Illumina MiSeq amplicon sequencing of the internal transcribed spacer (ITS) region, distance-decay relationships (DDRs), and ecological models, fungal assemblages and their spatial patterns were investigated from each soil depth. We observed significant spatial variation in fungal communities and found that environmental heterogeneity decreased with soil depth, while spatial variation in fungal communities showed the opposite trend. DDRs occurred only in 0-10cm- and 10-20cm-deep soil layers, not in the 20-40cm layer. Our analyses revealed that the fungal community assembly in the 0-10cm layer was primarily governed by environmental filtering and a high dispersal rate, while in the deeper layer (20-40cm), it was primarily governed by dispersal limitation with minimal environmental filtering. Both environmental filtering and dispersal limitation controlled fungal community assembly in the 10-20cm layer, with dispersal limitation playing the major role. Results demonstrate the decreasing importance of environmental filtering and an increase in the importance of dispersal limitation in structuring fungal communities from shallower to deeper soils. Effectively, "everything is everywhere, but the environment selects," although only in shallower soils that are easily accessible to dispersive fungal propagules. This work highlights that perceived drivers of fungal community assembly are dependent on sampling depth, suggesting that caution is required when interpreting diversity patterns from samples that integrate across depths.IMPORTANCE In this work, Illumina MiSeq amplicon sequencing of the ITS region was used to investigate the spatial variation and assembly mechanisms of fungal communities from different soil layers across paddy fields in subtropical China, and the results demonstrate the decreasing importance of environmental filtering and an increase in the importance of dispersal limitation in structuring fungal communities from shallower to deeper soils. Therefore, the results of this study highlight that perceived drivers of fungal community assembly are dependent on sampling depth and suggest that caution is required when interpreting diversity patterns from samples that integrate across depths. This is the first study focusing on assemblages of fungal communities in different soil layers on a relatively large scale, and we thus believe that this study is of great importance to researchers and readers in microbial ecology, especially in microbial biogeography, because the results can provide sampling guidance in future studies of microbial biogeography.
