Sequencing and soil fungal community composition
The PCR heminested approach produced a set of 3,524,761 raw sequences, with 353,312 high-quality ITS2 sequences, averaging 11,330 sequences per sample, retained after the denoising process. Twenty-six samples were used for downstream analyzes. Despite the higher number of valid sequences, land use MeA presented a lower number of OTUs. On the contrary, UnPA and AgA showed the highest number of OTUs (Table.1). Rarefaction curves suggest that a plateau was reached (Fig.2), indicating that the sampled environments have a good representation of the fungal community, with the identification of 489 OTUs, 9 phyla, 29 classes, 53 orders, 117 families, 158 genera and 174 species.
Among the 489 OTUs, 213 belonged to the Basidiomycota, 167 to the Ascomycota, 25 to the Mucoromycota, 17 to the Mortierellomycota, 16 to the Glomeromycota, and 41 to unassigned phyla. The heminested PCR approach allows identifying 10,529 (2.98%), high-quality ITS2 sequences of phylum Glomeromycota (4 orders and 6 families). Claroideoglomus (Family Claroideoglomeraceae) was the unique genus assigned. UnPA (5,302 sequences) has the highest number of Glomeromycota sequences, followed by MeA (2,709 sequences) and AgA (2,518 sequences) land uses. Some families such as Ambisporaceae and Acaulosporaceae were only present in the mining sites of La Ceja (LC-QC) (Table.2).
UnPA (40.95%) and MeA (56.29%) land uses have high relative abundance of Basidiomycota OTUs. However, the relative abundance of Ascomycota OTUs was the highest in AgA (32.40%) and displayed an increase in UnPA (37.20%). In addition, UnPA and AgA land uses showed an increase of Ascomycota, Mucoromycota, Glomeromycota and Mortierellomycota OTUs abundance, in comparison to MeA. The more abundant orders in the study zone were Hypocreales (16.26%), Agaricales (10.02%) and Malasseziales (8.88%). Hypocreales was the most abundant order in UnPA (3236 reads) and AgA (2466 reads). Malasseziales was the order with higher abundance in MeA (2611 reads). Nevertheless, this order was scarcely observed in UnPA (3 reads) (Fig.3).
UnPA contained 116 species, AgA 68 species and MeA 47 species. The most frequent species in every land use have similar prevalence (Table.3), with the frequent species in MeA land use being the less prevalent. Saitozyma sp. was the most frequent species in UnPA, whereas Trichoderma asperellum and Agrocybe pediades were the most frequent in AgA and MeA land uses. Malassezia sp. was prevalent and highly abundant in both perturbed land uses.
Soil fungal communities among land uses
The municipalities did not show significant differences at phylum and order rank. However, fungal taxa varied more strongly with land uses at the rank of order followed by the genera. Contrasting fungal genera abundances among land uses were only significant between UnPA-MeA (R2=0.139, p=0.002) and AgA-MeA (R2=0.121, p=0.01), with significant differences at the order rank in abundances between UnPA-AgA (R2=0.106, p=0.023), UnPA-MeA (R2=0.189, p=0.001) and AgA-MeA (R2=0.122, p=0.018). In addition, the abundance of dominant fungal orders displayed a meaningful distribution among land uses. For instance, samples from UnPA showed a high abundance of orders Hypocreales and Agaricales. Higher abundance of order Malasseziales was representative of MeA soil grouped samples. The cluster analysis of order composition shows two main groups of samples. The first group matches most of the samples retrieved on MeA while the second group involves samples from UnPA and AgA land uses (Fig.4).
Land uses show variations in the Alpha diversity indexes, with significant differences observed between UnPA-AgA (R2=0.412, p=0.006), UnPA-MeA (R2=0.709, p=0.001) and AgA-MeA (R2=0.514, p=0.002). However, the differences in abundances of taxa and alpha diversity indexes were not significant between municipalities. Higher alpha diversity values occurred in the UnPA land use, highlighting values observed in LC-MpMWF samples. Inversely, lower values occurred in MeA, especially in ER-QC and LC-QC (Fig.5). The alpha-diversity indexes showed consistent patterns among land uses (UnPA > AgA > MeA).
Fungal diversity responses to variations in soil properties and environmental factors
The studied Andosols have an acidic to weakly acidic pH range (from 5.3 to 5.7), with some parameters such as organic carbon content and soil moisture decreasing with the degree of perturbation. On the contrary, other parameters such as electrical conductivity and total dissolved solids increase with the degree of perturbation. Similarly, environmental parameters like relative air humidity and barometric pressure decrease in perturbed areas and other parameters such as soil temperature increase in those areas (Table.4). In AgA samples, pH and soil temperature decreased. In general, soil properties and environmental variables show a significant variation among land uses (R2=0.3182, p=0.0002).
The NMDS dissimilarity distance matrix with Bray-Curtis, performed to assess variability in community composition among land uses and specific activities, revealed that fungal communities of MeA were distinct from the other two land uses, observing a closer similarity between fungal communities from UnPA and AgA. The communities order composition positively correlates with relative air humidity and organic matter and negatively correlates with soil temperature and dew point temperature (Fig.6a). Three groups in the sample ordination, based on alpha diversity indexes, showed positive correlation with organic matter and negative correlation with pH (Fig.6b). The subset of indexes with the maximum rank correlation with samples dissimilarities was shaped by Shannon and Fisher index (r=0.94). Each vector had a substantial effect on the sample gradient ordination (p=0.001). However, every subset of indexes strongly correlates (r=>0.70) with dissimilarities in the communities.
Effect size measure of land use
UnPA and MeA showed a significantly high contrast in the ARR or effect size of differences observed in fungal community features and parameters assessed (Table.5). In general, land use intensity negatively affected abundances of prevalent fungal species, dominant order, alpha diversity index and environmental variables. The factors significantly affected by MeA and AgA land uses were the abundances of taxa belonging to order Wallemiales, Trichosporonales and organic matter. Despite this, the RRi soil moisture was significant in contrasting UnPA and MeA land uses, but not significant in contrasting UnPA and AgA land uses (Fig.7).