Research paperSelecting fungal disturbance indicators to compare forest soil profile re-construction regimes

- Fungal communities compared between undisturbed forest soils and disturbance regimes emulating soil reclamation practices.
- Disturbance indicators identified from amplified fungal ITS2 sequences using Linear Discriminant Effect Size algorithm.
- Exophiala, Hyphodontiella, Mastigobasidium, and Umbelopsis indicate disturbance across treatments and sampling depths.
- Ectomycorrhizal genera (e.g. Amphinema, Cortinarius, Piloderma, Russula) strongly indicate undisturbed soils.
- LEfSe comparison of stockpiled to immediately replaced LFH identified a single genus, Capronia, as differentially abundant.

Physical disturbance of soil profiles, even at shallow depths, is a ubiquitous consequence of anthropogenic landscape modification, with short-term impacts on important ecological guilds of fungi. DNA-based methods for surveying community composition are widely incorporated into studies attempting to explain fungal responses to forest ecosystem disturbances. Here we compare fungal community composition between three distinct soil profile manipulations (20 cm depth) and undisturbed control plots in a sub-boreal spruce forest in the Central Interior region of British Columbia, Canada. Fungal community composition differences were tracked by internal transcribed spacer 2 (ITS2) amplicon sequencing, with comparisons drawn using genus-level annotations. Non-metric multidimensional scaling (NMDS) analysis indicated that disturbed-sample community compositions were similar to controls at 0-months and distinct from controls at 5- and 12-months post disturbance, but did not indicate clustering of samples according to disturbance regime. We used Linear Discriminant Effect Size (LEfSe) analysis to identify fungal genera that consistently indicate disturbed or undisturbed (control) treatments across 5- and 12-month sampling times. Four fungal genera (Exophiala, Hyphodontiella, Mastigobasidium, and Umbelopsis) were detected with higher ranges of relative abundance in all disturbance regimes when compared to control plots. Deliberate mixing of LFH into mineral horizon soils stimulated multiple genera that were more frequently detectable in replicate plots at 12 months, when compared to undisturbed and immediately re-assembled plots. Four ectomycorrhizal genera (Amphinema, Cortinarius, Piloderma, and Russula) were identified as strong indicators of control plot soils. A single genus, Capronia, was identified as differentially abundant between stockpiled LFH and immediately replaced LFH. Our results are consistent with declines in ectomycorrhizal fungal abundance and increases in saprotroph abundance previously reported in DNA-based community profiling studies of forest soil disturbance. This investigation demonstrates that bulk soil sampling can be used to evaluate soil-handling regimens to understand fungal community disruption/recovery and highlights LEfSe as an approach to indicator selection in DNA-dependent biodiversity surveys.