Table 1.
Selection of fungal biocontrol candidates for ad planta trials with P. infestans on tomato plants. Data on isolates inhibiting A. solani, in vitro are included. The putative genera were determined by BLAST analysis after partial sequencing of the ITS locus (approximately 560 bp). Isolation parameters include media and temperature used for the isolation of fungi from plant samples. PDA: potato dextrose agar; OMA: oatmeal agar.
Fig 1.
Stepwise screening procedure of the fungal isolate collection for potential microbial antagonists (Modified from: Köhl et al. [32]).
Table 2.
Evaluation of the physiochemical parameters for soil A and soil B. The soil samples were analyzed according to VD-Lufa methods [59].
Fig 2.
Effect of P. infestans on tomato plants grown in two different soil origins (A and B), in terms of (A) plant fresh weight and (B) diseased leaf area (%).
The two soils were collected from two organic growers from Domäne Mechthildshausen, Wiesbaden (soil A) and Solidarische Landwirtschaft, Rüsselsheim (soil B). Plants were inoculated with P. infestans conidial suspension (2 × 104 conidia/ml) 5 weeks after sowing. Tomato plants inoculated with distilled water served as the healthy control. Four tomato plants were used for each treatment (n = 4). Data were collected 14 days after inoculation with P. infestans. Means were compared with two-way ANOVA (p ≤ 0.01) followed by Tukey’s post-hoc test (p ≤ 0.05). Significant differences are shown by letters (a, b).
Fig 3.
Activity of selected fungal isolates against P. infestans.
Dual-culture assays showing representative fungal candidates (right part of each Petri dish) and P. infestans (left part of each Petri dish). Assays were performed on PDA with both fungal plugs placed 4 cm apart on Petri dishes. The results were recorded after 7 days at 25°C for n = 3 plates.
Fig 4.
Efficacy of selected fungal isolates against P. infestans on tomato plants across three trials.
Spore suspensions (4 × 105 CFU/ml) prepared from Chaetomium sp. (Pf101), Trichoderma sp. (Pf131), Pseudogymnoascus sp. (Pf23), Chlonostachys sp. (Pf27), and Ctenomyces sp. (Pf45) were preventively applied on tomato leaves, 24 hours before inoculation with P. infestans (2 × 104 ml-1 conidia). Tomato plants inoculated with distilled water served as the negative control (Negative_control). Tomato plants inoculated with P. infestans only served as the positive control (Positive_control). Tomato plants treated with Cuprozin progress (0.52%) and inoculated with P. infestans served as the chemical control (Copper standard). Six tomato plants were used for each treatment (n = 6). Percentage disease leaf area (%) was determined 7 days after inoculation with P. infestans. The letters above the boxes show significant differences across treatments (p ≤ 0.05) according to Tukey HSD tests following ANOVA on diseased leaf area data. The three trials represent three independent replicates.
Fig 5.
The alpha-diversity indices species richness (A), Shannon diversity index (B), evenness (C), and Simpson index (D) of the fungal microbiome in samples from phyllosphere, endosphere, and rhizosphere from tomato plants grown in two soil origins (soil A and soil B).
Plants were grown in two different soil origins (A and B) and were inoculated with P. infestans (diseased, (turquoise plots) or remained untreated (healthy, red plots) 14 days before sampling was done. Four tomato plants were used for each treatment (n = 4). Values of alpha-diversity indices are shown, corresponding p-values are written above each comparison. Means were compared with a Kruskal Wallis test, with a significance threshold set at p ≤ 0.05.
Fig 6.
Microbial beta-diversity of different tomato micro-compartments visualized using an NMDS plot.
The fungal microbiome composition was determined using Bray–Curtis community dissimilarities. Plants were grown in two soil origins (A and B) and inoculated with P. infestans 14 d prior to sample collection from the rhizosphere, phyllosphere, and endosphere of healthy and diseased plants. ANOSIM was used to determine significant differences (p ≤ 0.001, S5 Table).
Fig 7.
Fungal taxonomic distribution of the microbiome analyses at the phylum, genus, and amplicon sequence variant (ASV) levels in the rhizosphere, phyllosphere, and endosphere of soil A and soil B.
The overlapping region represents taxa that are shared between soil A and soil B. Numbers inside each section represent the number of taxa at each level (phylum, genus, or ASV). The Venn diagram is based on the analysis of relative abundance with ANCOM-BC2.
Fig 8.
Diversity of the most abundant fungal taxa at phylum (A), genus (B), and (C) ASV levels between plants grown in soil A and soil B across the three micro-compartments (endosphere, phyllosphere, and rhizosphere).
Others represent all taxa with less than 9%, 6%, and 5.5% abundance at phylum, genus, and ASV levels, respectively. Bars represent the mean relative abundance (%) of each phylum genus, and ASV. Error bars indicate standard deviation. Asterisks indicate a significant difference in abundance between samples from soil origins A and soil B inferred at p-value ≤ 0.05 according to Benjamini-Hochberg (BH) correction, which controls for false discovery rate (FDR) analyzed using ANCOM-BC2 (Analysis of Compositions of Microbiomes with Bias Correction).
Table 3.
Taxonomic affiliation of the top most abundant fungal ASVs in the microbiome of tomato plants in the whole data set. The ASVs were taxonomically assigned using NCBI BLAST (megablast, v.2.14.1) with an e-value of 0.001 and the identity cut-off of 80% against the UNITE database (version 9, [46]) containing ITS sequences in a Galaxy workflow.