https://orcid.org/0000-0002-0627-0058
Figures
Abstract
Late blight is a disease whose causative agent is the oomycete Phytophthora infestans. It is one of the most destructive pathogenic oomycetes and a major challenge to global tomato production. The pathogen is difficult to manage because of its ability to evolve thereby evading host resistance. The aim of this study was to screen for potential antagonists of P. infestans using a combination of culture and microbiome-based approaches. Samples were collected from healthy and P. infestans-infected tomato plants grown in soil collected from two organic tomato growers in the Rhine-Main area in Germany. A total of 246 fungal isolates were screened for their antagonistic activity against P. infestans. Most of the isolates that exhibited in vitro antagonistic activity were from the genera Penicillium, Trichoderma, Chlonostachys, Mortierella, and Pseudogymnoascus. Following a stepwise in vitro screening strategy that accounted for growth features, ecological aspects, taxonomic data, potential health risks, commercial properties, and antagonistic efficacy, five fungal isolates were eventually selected for plant trials. Chaetomium subaffine showed the highest inhibitory effect against P. infestans across three trials whereby the percentage of diseased leaf area reduced by 90% compared to the control. Chlonostachys and Pseudogymnoascus spp. were effective in two trials, while Trichoderma and Ctenomyces spp. showed weak disease suppressive effects. In parallel, we characterized the fungal microbiome of the rhizosphere, phyllosphere, and endosphere from healthy and diseased tomato plants using ITS-rRNA sequencing. The fungal community differed significantly between the two soil origins, but P. infestans did not significantly influence fungal microbiota composition. Notably, 70% of our antagonistic fungi from the culture collection were detected in the tomato microbiome. This work identified isolates of Chaetomium subaffine, Clonostachys sp., and Pseudogymnoascus sp. as potential biocontrol candidates promoting plant health. The findings highlight the importance of combined functional screening and microbiome profiling for identifying fungal antagonists.
Citation: Orwa P, Kuhl-Nagel T, Meinhold-Ernst R, Seyer A, Jehle JA, Mwirichia R, et al. (2025) Chaetomium, Chlonostachys, and Pseudogymnoascus isolates from tomato tissues significantly suppress Phytophthora infestans in tomato. PLoS One 20(10): e0335007. https://doi.org/10.1371/journal.pone.0335007
Editor: Estibaliz Sansinenea, Benemérita Universidad Autónoma de Puebla: Benemerita Universidad Autonoma de Puebla, MEXICO
Received: June 11, 2025; Accepted: October 6, 2025; Published: October 24, 2025
Copyright: © 2025 Orwa et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Raw reads are deposited at the Sequence Read Archive (https://www.ncbi.nlm.nih.gov/sra) under the BioProject accession number PRJNA1273108. All scripts including detailed parameters and data to reproduce the figures for this study have been deposited in GitHub (https://github.com/TKuhl-Nagel/Phytophthora_challenged_tomato_microbiome.git).
Funding: P.O. received funding from Deutscher Akademischer Austauschdienst (DAAD) doctoral scholarship (award no. 91832749). The funders did not contribute to the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Tomato is an important crop forming a vital component of the global diet, owing to its nutritional and health benefits [1]. In addition to other fungal pathogens such as Alternaria solani, Botrytis cinerea, and Fusarium oxysporum f. sp. lycopersici [2], late blight disease caused by Phytophthora infestans, is among the most devastating pathogenic oomycetes threatening global tomato production [3]. Studies indicate that P. infestans has become more aggressive over the past decades [4,5] and that the rapidly evolving genotypes are linked to severe epidemic global outbreaks [6]. Management costs and yield loss associated with late blight disease ranges between 3–10 billion United States Dollars per year globally [7,8].
Conventional approaches for managing P. infestans primarily utilize synthetic fungicides, which adversely impact the environment and may harm terrestrial and aquatic organisms [9]. The pathogen also poses the risk of fungicide resistance as it evolves rapidly [10]. In this regard, there are increasingly stringent regulations regarding the use of chemical-synthetic and copper fungicides. The European Union (EU) for instance, has heightened restrictions on the registration of chemical-synthetic pesticides through directives 2009/128/EC and (EU) 2019/782, aiming at lowering the risks associated with pesticide application on humans and the environment [11]. Therefore, there is an urgent need for more eco-friendly alternatives to chemical-synthetic and copper fungicides. Microbial antagonists promote plant growth and suppress plant pathogens through distinct mechanisms, including induced systemic resistance (ISR) [12], competition for space and nutrients [13], or the production of antimicrobial secondary metabolites [14]. The genera Bacillus and Pseudomonas are commonly found as microbial antagonists [15,16], while Trichoderma is among the most investigated and described fungal genera used as a biocontrol agent [17,18].
The soil microbiome includes many beneficial microorganisms and is an important source for plants to recruit their rhizosphere microbiome [19]. Soil microbiomes provide unique properties for plant pathogen control, also known as soil-borne legacy, which is beneficial to plants of the next generation thriving on the same soil [20]. However, plant pathogens can drive dysbiosis of the plant microbiome with consequences for plant health [21]. The rhizosphere, shaped by plant roots, is a complex habitat for diverse microorganisms, such as protozoa, mycorrhizal fungi, nitrogen-fixing bacteria, plant growth-promoting rhizobacteria (PGPR), and mycoparasitic fungi [22–25]. Fungi play multifaceted roles in the soil, including nitrogen fixation, and synthesis of plant growth–promoting hormones, they are natural biological agents against root pathogens and promoters of plant resilience under drought stress [26–28]. Additionally, fungi are key players in stabilizing soil organic matter by breaking down plant residues [29]. In recent years, there has been a growing emphasis on harnessing the diversity of soil fungi as a sustainable method to improve the quality of soil and boost crop productivity —an emerging strategy holding significant promise for the future of agriculture [30]. Several fungal species have been described with biocontrol potential against P. infestans [31]. Research on new or better biocontrol agents often focuses mainly on in vitro confrontation assay results as the basis for the selection of promising candidates, a majority of which do not meet the requirements for commercial use. The current work seeks to address the above-described research limitations by finding answers to the following questions. (i) Are the micro-compartments of tomato plants a suitable source for isolating fungal strains with effective biocontrol potential against P. infestans, which also meet selection criteria for later product development and registration? (ii) Does the microbiome structure differ between healthy and P. infestans-infected tomato plants grown in two soil origins, and can specific fungal groups be identified as key contributors to disease suppression? Therefore, we applied an integrated approach in the search for fungal antagonists against P. infestans by linking culture-dependent characterization to microbiome analyses and assessing the results in the context of tomato plant health. In the culture-dependent approach, we isolated and characterized potential fungal antagonists from diseased and healthy tomato plant samples in a stepwise selection procedure previously described by Köhl et al. [32] to select promising candidates for ad planta trials against P. infestans. The selection procedure incorporates commercial aspects, including growth on low-cost growth media, massive spore production, and evaluating potential health risks, in addition to antagonistic efficacy. In the microbiome approach, we characterized and compared the fungal microbiome of the phyllosphere, the endosphere, and the rhizosphere of diseased and healthy tomato plants grown in soils from different origins using high-throughput sequencing (HTS). Linking culture-dependent work to the microbiome provides deeper insights to characterize beneficial fungi for biocontrol of tomato late blight and other tomato diseases.
Methods
Soil sampling, tomato plant growth conditions, and inoculation with P. infestans
Tomato plants (Solanum lycopersicum `Red Robin´ cultivar) were grown in two different natural soils that were previously used for organic tomato cultivation in the greenhouse. No chemical-synthetic pesticides and fertilizers were applied previously and during the experiments. As in both soils tomato plants were grown in the previous growing season, these soils are enriched with organisms that are supportive of tomato plant health. The soils were collected from two different professional tomato growers in the Rhine-Main area at Domäne Mechthildshausen, Wiesbaden, 50°02′ N, 8°19′ E (soil A), and Solidarische Landwirtschaft Rüsselsheim, 49°58′ N, 8°25′ E (soil B) in February 2022. The contact to the growers and the corresponding field sites (soil A and B) was provided by Christian Fetzer, employee from the Landesbetrieb Landwirtschaft Hessen, a governmental educational and advisory institution of the State of Hesse, subordinate to the Hessian Ministry of Agriculture and Environment, Viticulture, Forestry, Hunting, and Home Affairs (HMLU). The growers gave permission to collect soil from their greenhouses for this study. Soil samples were taken at five positions throughout the greenhouse area with a depth of 20 cm and a mixed sample from all samples was used as substrate for the experiments. Eight tomato seeds were sown in soils from both locations (soil A and B) in each pot (size 8 x 8 x 8.5 cm) within 24 h after soil collection. Four seedlings for each treatment and soil origin were transferred to individual pots 14 days after sowing. Plants were grown at a temperature range of 19−22°C for subsequent inoculation with P. infestans (isolate 606 of JKI culture collection) as described in Drenker et al. [33]. Briefly, four 5-weeks-old plants each grown in soil A and soil B were spray inoculated with 7 ml of 2 x 104 ml-1 zoospores per plant. Four plants grown in each soil were not inoculated and served as healthy controls. Scoring for disease symptoms (percentage symptomatic area) was performed 2 weeks after inoculation with P. infestans.
Sample preparation for microbiome analysis and isolation of fungal antagonists
Samples for partial ITS amplicon sequencing and isolation of viable fungi were collected from the above-described tomato plants. Samples from leaves, roots, and rhizosphere soil were collected from the tomato plant. One plant was considered one replicate, in total four replicate plants per treatment were analyzed. Several plant tissues per micro-compartment were pooled as described in detail below. Plant samples were processed for ITS amplicon sequencing and isolation of fungal antagonists following the protocol of Wieland et al. [34], with some modifications, described as follows.
Rhizosphere sample collection
The roots of each plant were shaken carefully to remove non-adhering soil. Five grams of soil adhering to the roots were transferred in sterile 50 mL centrifuge tubes containing 20 mL of sterile saline buffer (0.85% NaCl) and mixed vigorously by vortexing for 2 min. The roots (approximately 5 g), from which the rhizospheric soil had been dislodged were soaked in 20 mL of sterile saline buffer (0.85% NaCl) and mixed thoroughly by vortexing for 5 min to obtain the rhizoplane. Both solutions were combined to obtain the final rhizosphere sample. A total of 16 rhizosphere samples were obtained for microbiome analysis, four replicates from each soil origin, both from healthy and diseased plants. The remaining rhizosphere material from the four replicate plants was pooled for the isolation of fungal antagonists. A total of four rhizosphere samples were obtained for fungal isolation, one replicate from each soil origin, and from healthy and diseased plants.
Endosphere sample collection
Roots (approximately 5 g) from rhizoplane extraction described above were sterilized with ethanol (75%; 2 min), sodium hypochlorite solution (50%; 2 min), and ethanol (75%; 1 min) and rinsed five times in sterile distilled water. Root samples (100–150 mg) were then homogenized with a sterile pestle and mortar in 20 mL of sterile saline buffer (0.85% NaCl) to obtain the final endosphere sample. In total, 16 endosphere samples were obtained for microbiome analysis, four replicates from each soil origin, and healthy and diseased plants. The remaining endosphere material from the four replicate plants was pooled for the isolation of fungal antagonists. A total of four endosphere samples were obtained for fungal isolation, one replicate from each soil origin, and from healthy and diseased plants. A sterility check was performed by placing a subset of the sterilized roots on PDA at 27°C for 7 days, and the absence of microbial growth ensured the sterility of the root surfaces.
Phyllosphere sample collection
Samples for the microbiome analysis and isolation of leaf-associated fungi followed the protocol by Kassa Semagn [35]: Fresh leaf samples (100–150 mg) were harvested from each plant by pooling 4 portions obtained from different branches, cut into smaller pieces into the collection tubes and processed immediately. Notably, samples from plants infected with P. infestans were collected from leaf patches with healthy-looking samples adjacent to the diseased area. A total of 16 rhizosphere samples were obtained for microbiome analysis, four replicates from each soil origin, from both healthy and diseased plants. The remaining phyllosphere material from the four replicate plants was pooled for the isolation of fungal antagonists. A total of four phyllosphere samples were obtained for fungal isolation, one replicate from each soil origin, and from healthy and diseased plants.
Collection of fungal isolates
Culturable fungi were isolated from the phyllosphere and rhizosphere of healthy and diseased plants grown in soil A and B. A total of eight samples were collected for fungal isolation, including four rhizosphere samples (as described above) and four phyllosphere samples as described above. Precisely, 100 µl of serially diluted samples (10−1 − 10−3) in sterile saline buffer (0.85% NaCl) were spread-plated onto potato dextrose agar (PDA) and oatmeal agar (OMA) supplemented with antibiotics (rifampicin (20 mg/ml) and streptomycin (50 mg/ml)). Plates were incubated at different temperatures (14, 21, and 28°C) to increase the diversity of the collection. The first assessment of fungal growth was conducted 7 days after plating and was based on differences in growth morphology, whereas the last assessment of fungal growth was conducted 14 days after plating. Fungal colonies that emerged on plates with different media, at varying temperatures, and from different tissue samples were isolated through a series of sub-culturing steps.
Characterization of fungal isolates regarding their biocontrol potential
The present work adopted a stepwise screening approach developed by Köhl et al. [32] to select fungal candidates with promising characteristics as microbial antagonists of tomato pathogens (Fig 1). The isolated 246 fungi were examined under the microscope for their ability to form spores as this is a key parameter in the mass production of fungi for commercial purposes (S12 Table). The fungal isolates were further incubated at 37°C and those that grew in this temperature were eliminated as they could present potential risks to majorly humans and other mammals (S12 Table). Additionally, morphologically identical isolates that appeared repeatedly and those that lost viability were eliminated. The remaining 165 fungal isolates were subjected to dual-culture assays with P. infestans as the target plant pathogen in this work (S11 Table). A total of 61 fungal isolates that either formed inhibition zones when co-cultured with the pathogen on Petri dish or overgrew the pathogen were deemed to exhibit antagonistic effects, through antibiosis and matrix completion, respectively (S9 Table). The fungal antagonists were identified as potential biocontrol candidates and were molecularly characterized by sequencing the ITS locus, yielding an amplicon of approximately 560 bp (S9 Table). The already identified antagonists were further screened against A. solani (isolate 62028 of JKI culture collection) to reveal their broader antagonistic effects (S9 Table). Notably, the in vitro confrontation assays with A. solani only served as criteria for selecting the most promising antagonists, therefore, the pathogen was not used in the subsequent experiments. A literature search was performed and species that have previously been described as plant, human, or animal pathogens were eliminated, particularly Penicillium species which formed a higher proportion of our collection. The resultant collection of 24 fungal isolates were re-screened for in vitro inhibitory activity on P. infestans but this time on rye agar, which is the most suitable growth media for the pathogen (S10 Table). In the end, 5 fungal antagonistic candidates were selected for ad planta tests with P. infestans on tomato plants (Table 1). The criteria for selection of the 5 potential fungal antagonists included (i) literature evidence as a potential plant biocontrol agent: both broadly described genera, including Trichoderma and scarcely described genera, including Pseudogymnoascus were considered; (ii) ability to produce at least 4 x 105 ml-1 spores on standard fungal media (PDA) within 21 days of incubation at 25 ˚C; (iii) no negative effects on healthy tomato plants following leaf application.
In vitro dual-culture assays of fungal isolates against P. infestans and A. solani
Two mycelial discs (5 mm) of 7-day-old agar cultures of the fungal isolates grown on PDA and 7-day-old cultures of P. infestans grown on rye agar (200 g rye; 3 g glucose; 12 g agar; 1000 ml distilled water) were simultaneously inoculated 4 cm opposite each other on PDA in the first screening of 165 isolates (S11 Table) and on rye agar in the second screening of 24 isolates (S10 Table) in standard Petri dishes. The control plates were inoculated with P. infestans only. Three Petri dishes were prepared for each isolate. The plates were incubated at 25°C for 6–7 days, and fungal pathogen growth reduction indicated by inhibition zones around P. infestans was recorded (yes- inhibition zone present and no- inhibition zone absent). Fungal isolates that grew over the pathogen were also recorded as (yes) showing potential inhibitory effect. The screening for A. solani, precultured on PDA medium was performed in the same way as described for P. infestans.
Testing disease suppressive effects of selected fungal isolates ad planta
Five fungal isolates, Chaetomium sp. (Pf101), Trichoderma sp. (Pf131), Pseudogymnoascus sp. (Pf23) Chlonostachys sp. (Pf27), and Ctenomyces sp. (Pf45), were selected from the characterized fungal collection for plant trials (Fig 1). The isolates were grown on PDA for 21 days at 25°C for spore production. Tomato seeds (`Red Robin´, Weigelt Samen) were sown in a 3:1 mixture of substrate (ProLine Potgrond, Klasmann-Deilmann, Germany) and sand, respectively, in the greenhouse. The seedlings were transplanted into planting pots (8 x 8 x 8.5 cm) containing the same soil mixture, 14 days after sowing. The plants were maintained in the greenhouse with demand-oriented watering and weekly fertilizer application using 0.2% (w/v) Hakaphos blau (Compo Expert, Münster, Germany). From each of the above-mentioned fungal candidates, a spore suspension of 4 x 105 ml-1 was adjusted in distilled water. Five ml were sprayed on the leaves of 5-week-old plants. Plants were kept in the dark overnight. On the next day, the plants were spray-inoculated with 5 ml 2 x 104 ml-1 conidia of P. infestans. The trays were covered with lids to ensure 100% humidity and left in the dark overnight. The plants were placed in the growth chamber (Grow Bank, Plant Climatics, Germany) set to 16 h/8 h light/darkness, day temperature 21°C, night temperature 16°C, day humidity 65%, and night humidity 75%. Percentage diseased leaf area was visually recorded 7 days after spraying with fungal biocontrol candidates as described by Drenker et al. [33]. Six tomato plants were used per treatment, and three independent replicates of the plant trials were done. All plant trials followed a randomized design.
Molecular identification of isolates with inhibitory effects against P. infestans
Fungal DNA was extracted from 7-day-old cultures grown on PDA at 28°C using the DNeasy Plant Mini kit (Qiagen, Germany) following the manufacturer’s protocol. DNA quality was determined spectrophotometrically (NanoDrop 2000c; peqLab Biotechnologie GmbH, Erlangen Germany). The Internal Transcribed Spacer (ITS) region (approximately 560 bp) was amplified using primers ITS 1F- 5’-TCCGTAGGTGAACCTGCGG-3’, and ITS 4–5’-TCCTCCGCTTATTGATATGC-3’, as described in White et al. [36]. Briefly, a 25 µl PCR reaction mix was prepared for each sample, with reagents quantified as outlined by the manufacturer (Axon Labortechnik, Kaiserslautern). In the cases when PCR amplification was not successful, PCR reagents with a GoTaq® G2 Flexi DNA polymerase from Promega were used. The PCR cycling parameters were as follows: initial denaturation at 95°C for 3 min; 40 cycles at 95°C for 30 s (denaturation), 56°C for 30 s (annealing), and 72°C for 1 min (extension); final extension at 72°C for 5 min. PCR products were purified using the ExoSAP-IT™ PCR Product Cleanup Reagent (Thermo Fisher Scientific) and 10 µl was used for sequencing (StarSEQ, Mainz), pre-mixed with ITS-1F. The Basic Local Alignment Search Tool (BLAST) was used to identify potential genus matches by comparing the sequences of the isolates against the National Center for Biotechnology Information (NCBI) database (blastn, [37], and the most relevant hits based on the highest query coverage and percentage identity were recorded.
Molecular identification of Chaetomium species
The actual species of the Chaetomium isolate was evaluated by sequencing three identification loci according to Linkies et al. [38]. Internal transcribed spacer (primers ITS-1 and ITS-4 [36]; RNA polymerase II second largest subunit (primers RPB2AM-1bf (5’-CCA AGG TBT TYG TSA ACG G-3’) and RPB2AM-7R (5’-GAA TRT TGG CCA TGG TRT CCA T-3’) [39]; β-tubulin (primers T1 (5’-AACATGCGTGAGATTGTAAGT-3’) and T22 (5’-TCTGGATGTTGTTGGGAATCC-3’) [40] were employed with the following PCR conditions: Initial denaturation: 5 min at 95°C, 40 cycles of denaturation: 30 s at 95°C, 30 s of annealing at 55°C and 45 s of extension at 72°C, and a final extension of 5 min at 72°C. The PCR product was visualized by 1% agarose gel electrophoresis. Cleaned PCR products using the ExoSAP-IT PCR Product Cleanup Reagent (Thermo Fisher, Berlin, Germany) were sequenced at StarSEQ (Mainz, Germany). Sequencing was performed in one direction and the species name was recorded as the highest BLAST hit. A consensus sequence of three separate PCR reactions was examined. The top BLAST hits for all three examined coding regions showed a 100% match to Chaetomium subaffine in both the ITS and β-tubulin regions, and a 99.63% match in the RNA polymerase II second largest subunit region.
DNA extraction and library preparation for the fungal microbiome analysis
Total DNA was extracted from 0.25 mg of samples of the phyllosphere, endosphere, and rhizosphere from both soil origins and both healthy and diseased plants in four replicates (48 samples in total) using the DNeasy Powersoil Pro kit (Qiagen, Germany) following the manufacturer’s protocol. DNA yield and quality were checked photometrically (NanoDrop 2000c; peqLab Biotechnologie GmbH, Erlangen Germany) and samples were stored at −20°C. Library construction and high-throughput amplicon sequencing of ITS region with primers ITS1F (5’-CTTGGTCATTTAGAGGAAGTAA-3’ [41] and ITS2 (5’-GCTGCGTTCTTCATCGATGC-3’ [36] was carried out by StarSEQ GmbH (Mainz, Germany). Illumina adapter sequences were added in a 1-step PCR approach. Libraries were normalized and sequenced on the Illumina MiSeq platform with a v3 Reagent Kit in 2 x 300 nt paired-end mode.
Analysis of ITS rDNA amplicon sequencing
Primers and adapters were removed from high-quality raw reads using the software Cutadapt (v.4.4, [42], implemented in the European Galaxy-Server [43](https://usegalaxy.eu/). Processing of paired-end reads was performed in R (v.4.2.1, [44] using the DADA2 pipeline (v.1.24.0, [45] following the adapted workflow for ITS sequences (DADA2 ITS Pipeline Workflow (1.8), https://benjjneb.github.io/dada2/ITS_workflow.html, accessed 25.1.2024). The obtained amplicon sequence variants (ASVs) were taxonomically assigned using NCBI BLAST (megablast, v.2.14.1, [37] with e-value 0.001 and the identity cut-off of 80% against the UNITE database (version 9, [46] containing ITS sequences in a Galaxy workflow as described in Fernandez-Grecco et al. [47]. Sequences identified as not fungal (= other eukaryotes) and sequences with less than five reads were removed from the dataset. This resulted in a final number of 2,124 ASVs and on average 124,032 quality reads per sample from 3,962,154 raw reads. Rarefaction curves indicated sufficient sequencing depth (S1 Fig). Raw reads are deposited at the Sequence Read Archive (https://www.ncbi.nlm.nih.gov/sra) under the BioProject accession number PRJNA1273108.
All scripts including detailed parameters and data to reproduce the figures for this study have been deposited in GitHub (https://github.com/TKuhl-Nagel/Phytophthora_challenged_tomato_microbiome.git).
Data analysis and statistics
Statistical analysis and data handling for ASVs was performed using R (v.4.2.1, [44]) with the R packages “tidyverse” (v.1.3.2, [48], “ggplot2” (v.3.5.1, [49]), “ggtext” (v.0.1.2, [50]), “dplyr” (v.1.1.4, [51]), “tibble” (v.3.2.1, [52]), “stringr” (v.1.5.1, [53]), “ARTool” (v.0.11.1, [54]), “vegan” (v.2.6–6.1, [55]) and “phyloseq” (v.1.44.0, [56]). To account for uneven sequencing depth, the dataset was rarefied by 1000 times random subsampling to the lowest number of reads (18,628 reads) (S1 Fig) and all further analyses were performed with the rarefied dataset as recommended by Schloss [57]. Significant effects of micro-compartment, soil origin, and disease status on alpha-diversity indices (Richness, Evenness, Shannon, Simpson) were calculated using the non-parametric aligned rank ANOVA because residuals failed the normality criteria based on the Shapiro test in the previously performed 3-way-ANOVA (S1-S4 Tables). After normalizing the dataset to relative abundance (%) Bray-Curtis distance was used to estimate beta-diversity and visualized using nonmetric multidimensional scaling (NMDS). The effects of micro-compartment, soil origin, and disease status on the fungal community composition were tested using PERMANOVA (10,000 permutations). Relative abundance was computed for each ASV within each sample by dividing ASV counts by the total count per sample. Square bar plots were generated to represent the most abundant taxa (phylum, genus, and ASV level) based on their relative abundance in the whole dataset. Differences in their relative abundances based on soil origin were calculated for each micro-compartment using ANCOM-BC2 (v2.1.2, [58]) with p-value correction via the Benjamini-Hochberg method. The data set was filtered for rows with 0 ASVs prior to ANCOM-BC2 (Analysis of Compositions of Microbiomes with Bias Correction) to reveal differentially abundant taxa considering the two soil origins. ANCOM-BC2 also calculates the bias-corrected log-fold change (LFC) for each taxon between the two soil conditions to confirm that any taxon found only in one condition and completely absent in the other is a structural zero (sz) rather than rarely detected low-abundant taxa. LFC represents the ratio of the abundance of taxon between two conditions on a logarithmic scale. Negative LFC values indicate higher abundance in Soil A, while positive values indicate higher abundance in Soil B. ASVs classified as structural zero are exclusively present in one soil origin and are represented by strongly negative LFC (approaching -∞) and strongly positive LFC (approaching +∞). In order to compare the isolated living fungal species to the fungal community of the rhizosphere, we mapped the ASV sequences (~250 bp) to the isolate sequences (~500 bp) with 100% identity and at least 50% query coverage using NCBI BLAST. Afterwards, isolates were evaluated for their presence among the most abundant taxa and/or the differentially abundant taxa in the microbial community. Statistics for ad planta experiments and generation of plots were performed in R (versiolastn 4.4.2). Data on diseased leaf area (%) were presented as mean values ± standard error (SEM) computed from 6 biological replicates. The ad planta inhibitory effect of the fungal candidates against P. infestans was evaluated using a two-way analysis of variance (ANOVA) as the data was normally distributed according to Shapiro-Wilk test results. Significant differences between means were evaluated by Tukey’s test at P ≤ 0.05. Plant experiments were conducted in three independent repetitions, analyzed separately. Raw data from the three plant trials, and the associated R scripts including detailed parameters and data to reproduce the figures for this study have been deposited in GitHub (https://github.com/TKuhl-Nagel/Phytophthora_challenged_tomato_microbiome.git).
Results
Comparison of soil properties from Wiesbaden and Rüsselsheim in the the Rhine-Main area of Germany
Both soils were loam soils with comparable measures in terms of physical properties including humus level and organic carbon content (Table 2). Soil A had a slightly higher pH than soil B. While both soils were sufficiently supplied with macronutrients such as phosphorus, potassium, magnesium, and trace elements, they had low iron content. Tomato plants grown in soil A demonstrated a significant difference in weight between healthy and diseased conditions following 14 days of infection with P. infestans (Fig 2A), with diseased plants showing a significantly lower fresh weight. In soil B, there was no significant difference in fresh weight between healthy and diseased plants. However, healthy plants grown in both soil origins had comparable measures in weight. In addition, plants grown in both soils showed slight differences in the degree of symptom development upon inoculation with P. infestans (Fig 2B). The results indicate that the selected soils have comparable properties and are suitable as a resource for plant-beneficial fungi.
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).
Characterization of the fungal culture collection derived from tomato rhizosphere and phyllosphere
The isolation process of fungi from different samples yielded 246 isolates (Fig 1). In a stepwise screening approach described in the methods section, the number was reduced to 61 isolates (S9 Table) with promising characteristics as biocontrol agents against P. infestans. The in vitro antagonistic activity of the fungal isolates against P. infestans served as one selection criterion (Fig 3). Additionally, parameters that are advantageous for later product development and approval or registration as a biocontrol product were also taken into account, namely the inability to grow at 37°C, easy spore induction, and ability to survive on standard growth media. Out of the 61 identified isolates with antagonistic effects against P. infestans, the rhizosphere yielded 45 fungal isolates, of which 24 and 21 isolates derived from healthy and diseased rhizosphere, respectively. Another 16 isolates were recovered from the phyllosphere, of which eleven isolates derived from diseased phyllosphere compared to five isolates from healthy phyllosphere. All candidates were identified on the genus level by partial sequencing of the ITS locus.
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.
The genus Penicillium constituted the largest share (27 isolates), followed by Trichoderma (6 isolates), Agrostalagmus (4 isolates), Chlonostachys, Mortierella, and Pseudogymnoascus (3 isolates each), (S9 Table). All other genera were only present once or twice. Based on literature research, potential health concerns and plant pathogenic isolates, such as Penicillium, Fusarium, and Acrostalagmus were removed, ending up with 24 remaining candidates (Table 1). Among the 24 fungi, Mortierella sp., Pseudogymnoascus sp., Trametes sp., and Trichoderma sp. are notable for their potential to inhibit both P. infestans and A. solani. Among the antagonists tested for ad planta disease suppressive effects on P. infestans included Pseudogymnoascus sp. and Trichoderma sp.. However, no further tests were performed with A. solani as most of the antagonists showed weak inhibition potential via in vitro confrontation assays. Intriguingly, a majority of the selected fungi (17 isolates) originated from samples of tomato plants grown in soil A.
Fungal isolates show disease-suppressive effects ad planta against P. infestans
Five of the remaining 24 fungal candidates were selected for exemplary tomato plant trials evaluating suppressive characteristics against P. infestans. Pseudogymnoascus sp. (isolate Pf23), Chlonostachys sp. (Pf27), and Ctenomyces sp. (Pf45), Chaetomium sp. (Pf101), and Trichoderma sp. (Pf131), were chosen based on visible mycelial growth towards the edges of the Petri dish on standard media, abundant spore production (at least 4 x 105 ml-1 spores in standard fungal media (PDA)), and literature research. The isolates demonstrated varying levels of disease-suppressive potential in growth chamber plant trials (Fig 4). Chaetomium sp. (Pf101) significantly suppressed P. infestans across three independent replicates of the trial, as evidenced by the lower percentage of diseased leaf area compared to the positive control. Chlonostachys sp. (Pf27) significantly suppressed P. infestans in the first and second trials, whereas Pseudogymnoascus sp. (Pf23) significantly suppressed P. infestans in the second and third trials. Ctenomyces sp. (Pf45) was able to inhibit P. infestans once and Trichoderma sp. (Pf131) did not significantly inhibit symptoms of P. infestans, although the percentage of diseased leaf area was slightly lower compared to the positive control.
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.
Soil origin influences the fungal community in tomato plants
The micro-compartment (endosphere, phyllosphere, rhizosphere), soil origin, disease status, and their interaction significantly influenced the alpha diversity of the fungal microbiome (ANOVA, Fig 5A-D, S1-S4 Tables). Rhizosphere samples had the highest alpha-diversity indices indicating the presence of many species and with a rather homogeneous distribution compared to endosphere and phyllosphere. The disease status of the plant had a significant influence on the alpha-diversity only in the phyllosphere (ANOVA, S1-S4 Tables), which shows significant differences in the within-sample diversity between healthy and diseased plants mainly in soil B (Fig 5A-C). Besides, P. infestans had an effect on fungal species richness in the endosphere of plants grown in soil A, where infection reduced fungal species richness compared to healthy samples (Fig 5A). These results show that leaf infection with P. infestans led to changes in the within-sample microbiome composition of the phyllosphere, although these changes are only significant in plants grown in soil B. In contrast, the endosphere and rhizosphere were mainly not significantly affected by microbial diversity after inoculation with P. infestans.
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.
The micro-compartment, soil origin, and their interaction were identified as the driving factors of the fungal ß-diversity (PERMANOVA p ≤ 0.001, S5 Table, Fig 6), whereas P. infestans infection had no significant influence on the fungal community composition. Phyllosphere samples clustered more distant from each other compared to rhizosphere and endosphere.
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).
Based on these results we focused on the effect of soil origin on the fungal microbial community in the different micro-compartments because beta diversity analysis showed no significant difference between diseased and healthy plants. In each micro-compartment, we identified shared and unique taxa between the two soils (Fig 7), using ANCOM-BC2 (Analysis of Compositions of Microbiomes with Bias Correction). Rhizosphere samples of the two soil origins had overall more taxa compared to the phyllosphere and the endosphere, which confirms previous results. Interestingly, samples from soil B harbour more taxa compared to soil A across the three micro-compartments, indicating a soil-specific effect on fungal composition at the phylum, genus, and amplicon sequence variant (ASV) levels. At the same time, a substantial number of taxa were shared between plant samples from soil A and soil B across the three taxonomic levels (phylum, genus, ASV).
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.
Further analysis of the most abundant taxa in plant samples from the two soil origins showed that the phylum Ascomycota had the highest relative abundance in the whole data set (Fig 8A). Although other phyla showed variation between the two soil origins, only Basidiomycota, Rozellomycota, and Mortierellomycota significantly differed in abundance in individual micro-compartments. At the genus level, Botryotrichum, Geosmithia, GS11_gen_Incertae_sedis, and Pseudopyrenochaeta were more abundant in soil A, whereas Aspergillus, Brunneomyces, and Verticillium were more abundant in soil B in one or more micro-compartments (Fig 8B). Cladosporium and Acremonium were the top abundant genera in the phyllosphere, while Colletotrichum and Pseudopyrenochaeta were the most abundant group in the endosphere. The rhizosphere had a high proportion for an unknown genus, GS11_gen_Incertae_sedis.
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).
The most abundant ASVs in the whole dataset were from three phyla, with a majority from Ascomycota, followed by Olpidiomycota, and Rozellomycota (Fig 8C, Table 3). The differentially abundant ASVs between the two soil origins across the three micro-compartments were linked to Cladosporium, Geosmithia, Olpidium, Pseudopyrenochaeta, Verticillium, and unknown genus (GS11_gen_Incertae_sedis) (Fig 8C, Table 3).
Comparison of culture collection and fungal microbiome composition
In order to identify overlaps between the culture collection of living isolates and the microbiome analysis, we mapped ITS sequences of selected fungal isolates that demonstrated in vitro inhibitory effects against P. infestans to fungal ASVs in the microbiome of tomato plants grown in soils A and B. Of the 61 fungal isolates, 42 (Penicillium (21), Trichoderma (5), Acrostalagmus (4), Clonostachys (2), Pseudogymnoascus (2), Cladosporium (2), Bionectriaceae (2), Ctenomyces (2), Chaetomium (1) and Humicola (1)) showed 100% identity with the ASV sequences.
Cladosporium linked to isolates Pf179 and Pf242 from our culture collection was the only genus detected among the top 15 ASVs in terms of relative abundance (Table 3). Of the ASVs that matched our fungal collection, only ASV14 identified as Cladosporium sp. was differentially abundant in the rhizosphere between the two soil origins (Fig 8C, Table 3), with higher abundance in soil A. The matching ASVs were further evaluated for exclusive presence (also referred to as structural zeros) in plant samples from either soil origin. Intriguingly, ASVs identified as the genera Mortierella, Ctenomyces, and Pseudogymnoascus were exclusively present in the rhizosphere of plant samples from soil A, whereas ASVs identified as the genus Chaetomium were exclusively found in samples of plants grown in soil B (S6 Table). Phyllosphere and endosphere samples showed a similar pattern of exclusive taxonomic presence, with plants cultivated in soil A enriched with ASVs identified as the genus Ctenomyces and those from plants grown in soil B enriched solely with ASVs identified as Chaetomium (S7-S8 Table).
Discussion
In times of increasing awareness of sustainable use of pesticides, microorganisms gain more and more importance as biocontrol agents of plant diseases [60,61]. Our study characterized cultivable fungi from various plant micro-compartments as the initial step towards discovering fungal biocontrol agents against P. infestans. Additionally, we employed a fungal microbiome analysis to evaluate the potential importance of our fungal candidates within the microbiome. The initial number of 246 isolates obtained from the phyllosphere, endosphere, and rhizosphere of tomato plants was reduced to 24 potential biocontrol candidates remaining for plant trials by utilization of the step-wise screening procedure described by Köhl et al. [32]. The multi-step framework employs several steps to screen a high number of isolates for candidates fulfilling requirements for antagonists in commercial use, for example regarding efficacy, ecology, production, safety, and environmental risks. Schisler and Slininger [62] also included the above commercial parameters to select viable bacterial antagonists against Gibberella pulicaris, a potato dry rot causing pathogen.
The rhizosphere yielded more candidate isolates than the phyllosphere and endosphere, which concurs with previous findings that described the rhizosphere as a highly competitive habitat and a stable reservoir for beneficial microbes in terms of pathogen suppression, nutrient acquisition, and host defense functions [24,63]. Additionally, as a nutrient-rich environment, partly caused by plant root exudates [64], the rhizosphere supports a diverse microbial community, including beneficial fungi with potential biocontrol characteristics. Low microbe numbers recovered from the phyllosphere are linked to the specific nature of this habitat which is characterized by highly varying conditions that support only the survival of a limited number of microbes [65]. Recovering more fungal candidates from diseased phyllosphere samples suggests the enrichment of fungal microbial antagonists in pathogen-prevalent micro-compartments. Phyllosphere-associated microbes have been described with various functions, including plant growth promotion and pathogen inhibition [66], and nitrogen fixation [67], and can be causative agents for plant diseases [66]. The genera Penicillium, Trichoderma, Chlonostachys, Mortierella, and Pseudogymnoascus dominated our collection of potential microbial antagonists based on in vitro tests. A majority of these fungal groups have been adequately described as potential biocontrol agents. Culture filtrates of Penicillium griseofulvum demonstrated inhibitory effects on the growth of P. infestans in vitro [68]. But since Penicillium [69] species have been frequently reported as mycotoxin producers or as otherwise harmful in regard to food safety and consumer health [70], all Penicillium sp. isolates were eliminated from subsequent steps at an early screening stage. Trichoderma has been broadly investigated as a BCA owing to its ability to thrive in diverse soil habitats and outcompete pathogens for space and nutrients among other biocontrol mechanisms [71]. Trichoderma is also linked to other plant-beneficial functions such as phosphate solubilization and could be vital for the overall fitness of tomato plants through plant growth promotion [72]. Particular strains of Chlonostachys rosea [73], a tomato stem endophyte, have been reported to significantly reduce early blight caused by A. solani [74] and B. cinerea [75] in tomatoes. Products based on Trichoderma sp. [76] and C. rosea [77] are approved as fungicides in Europe. Moreover, there is limited information on the isolation of Pseudogymnoascus species from tomato plants, but this fungal genus is among the most abundant taxa in tomato root microbiome through amplicon sequencing [78].
While in vitro effects of certain microbial species or strains may not necessarily be reflected in the in vivo effects, the five selected fungal candidates tested in ad planta assays showed varying disease-suppressive efficiency against P. infestans. The candidate antagonists from Chaetomium sp., Chlonostachys sp., and Pseudogymnoascus sp. demonstrated significant disease-suppressive effects in at least two separate plant trials, validating the feasibility of our approach in identifying new or improved biocontrol agents. It is worth noting that the findings on Pseudogymnoascus contribute substantially to the knowledge of this genus in the field of plant pathology as a potential biocontrol agent. Chaetomium species, including Chaetomium cochliodes and Chaetomium aureum, are known to produce antifungal secondary metabolites against tomato pathogens such as A. solani, B. cinerea, and P. infestans [38,79]. However, other species of Chaetomium induce human pathogenic effects linked to systemic infections and allergic reactions [80]. Our isolate was confirmed to be Chaetomium subaffine after sequencing three identification loci (as described in the methods). To the best of our knowledge, there is currently no data on potential health or environmental risks associated with C. subaffine. Liu et al. [81] revealed the potential mycolytic effects of C. subaffine LB-1 on phytopathogenic fungi B. cinerea and A. solani by altering their cell membrane and cell wall. Reduced biocontrol efficacy by other groups, including Trichoderma sp. (Pf131) and Ctenomyces sp. (Pf45) across trials could be linked to environmental or procedural effects. For instance, soil-originated beneficial microbes such as Trichoderma when introduced to the leaves and vice versa may not survive or could cause negative interactions with the native community of microorganisms on the leaf reducing the efficacy [3]. Further processing steps such as drying or adding stabilizers might surely change the outcome but have not been investigated here. Different application methods such as the use of soil additives and fertilizers or a regular leaf application allowing for preventive and curative effects might also increase efficacy [82] but have not been tested yet.
High-resolution microbiome profiling techniques have improved the research on cultivable fungal species with plant-beneficial effects, providing crucial data for the targeted application of antagonists to ensure improved efficacy [83,84]. One example of this approach is demonstrated by Sahu et al. [85] who integrated culture methods and metabarcoding to profile the microbiome of the rice phyllosphere, revealing the key microbial communities for blast pathogen control. Similarly, we detected the presence of some of our fungal candidate antagonists for P. infestans, including Cladosporium among the highly abundant taxa and differential taxa in the tomato plant microbiome dataset. Although some Cladosporium species are plant pathogens [86,87], there are plant-beneficial species promoting growth [88,89]. This genus could reflect the core microbiome which has received a lot of interest in recent years, with multiple studies revealing core microorganisms, including bacteria, fungi, and archaea playing different roles in various plant hosts [90–92]. The dominance of specific fungal taxa, for instance, the exclusive presence of Mortierella, Ctenomyces, and Pseudogymnoascus in the rhizosphere samples from soil A and Chaetomium from rhizosphere samples in soil B may be a reflection of their beneficial activities to ensure a healthy plant microbiome [83]. In the rhizosphere, root exudates released due to pathogen invasion trigger plants to recruit microorganisms that are involved in specific functions, including direct suppression of the pathogens [93,94]. The endosphere represents the internal phytomicrobiome enriched with specifically adapted microorganisms that interact intimately [95]. The phyllosphere is linked to the atmosphere and interacts continuously with the air microbiome [96]. The fact that the fungal candidate antagonists are naturally present within the tomato plant and are effective against P. infestans highlights their potential as BCAs [97,98]. Therefore, knowledge of the ecological role of fungal biocontrol candidates in their native microbiomes can provide insights into the selection of improved and effective fungal agents against P. infestans.
Enhanced knowledge of how pathogen infection of a host plant alters soil microbial communities may pave the way for the use of microbial antagonists in the long-term management of plant diseases [99]. Therefore, our approach to microbiome characterization provides a more complete picture of how P. infestans and distinct soil sources may influence the composition of the tomato fungal community. Although P. infestans did not alter the overall fungal microbiome composition of tomato plants, significant changes in fungal alpha diversity were evident in specific micro-compartments, particularly the phyllosphere. Previous research on P. parasitica (closely related to P. infestans) revealed changes in the bacterial microbiota in tomato roots upon infection, whereas the fungal microbiota remained relatively stable [100]. These changes in fungal composition could also be attributed to distinct soil physiochemical parameters as we cultivated tomatoes in soils from two different locations in the Rhine-Main area [101]. Studies have shown that pH plays a key role in determining the composition of bacteria and fungi in the rhizosphere and bulk soil [102,103]. Also, soils with sufficient nutrient amounts provide optimal habitat for beneficial microbial communities to outperform pathogens [104].
Taken together, we present a strategic approach that enhances the likelihood of identifying fungal candidates with both strong biocontrol potential and commercial viability. The tested fungal biocontrol candidates, including Chaetomium sp., Chlonostachys sp., and Pseudogymnoascus sp. isolated from different micro-compartments of tomato plants grown in two soil origins inhibited P. infestans and A.solani in vitro and reduced P. infestans-related disease symptoms ad planta. We have also demonstrated that P. infestans infection of tomato plants influences the fungal community composition at the micro-compartment level, particularly on the phyllosphere rather than on the overall plant microbiome. However, the effect of soil origin on fungal community composition is significant considering the overall microbiome of tomato plants. The high abundance and differential abundance of specific genera, in our case Cladosporium indicates the potential beneficial role of this genus in tomato plant health as initial treatment showed no negative effects on the plant. In subsequent steps, we plan to test the suppressive effects of our Chaetomium subaffine in open-field plant trials. Other fungal candidates with variable efficacy against P. infestans will be subjected to optimization steps and used as consortia. Lastly, we intend to incorporate both curative and preventive application methods in future plant trials to achieve better disease-suppressive effects from the fungal candidates in a controlled growth room and greenhouse conditions. These findings improve our knowledge of the link between fungal biological control agents and plant microbiome, providing valuable ecological data in identifying effective antagonists suited for sustainable control of P. infestans.
Supporting information
S1 Table. Non-parametric aligned-rank ANOVA results for species richness using the ARTool package (v.0.11.1).
https://doi.org/10.1371/journal.pone.0335007.s001
(DOCX)
S2 Table. Non-parametric aligned-rank ANOVA results for Shannon using the ARTool package (v.0.11.1).
https://doi.org/10.1371/journal.pone.0335007.s002
(DOCX)
S3 Table. Non-parametric aligned-rank ANOVA results for species evenness index using the ARTool package (v.0.11.1).
https://doi.org/10.1371/journal.pone.0335007.s003
(DOCX)
S4 Table. Non-parametric aligned-rank ANOVA results for the Simpson index using the ARTool package (v.0.11.1).
https://doi.org/10.1371/journal.pone.0335007.s004
(DOCX)
S5 Table. Permutation test results from Adonis2 for beta diversity analysis with Bray-Curtis dissimilarity.
https://doi.org/10.1371/journal.pone.0335007.s005
(DOCX)
S6 Table. ANCOM-BC2 results of differential abundance of fungal ASVs between soil A and soil B in rhizosphere samples based on Log Fold Change (LogFC) and structural zero analysis.
https://doi.org/10.1371/journal.pone.0335007.s006
(DOCX)
S7 Table. ANCOM-BC2 results of differential abundance of fungal ASVs between soil A and soil B in phyllosphere samples based on Log Fold Change (LogFC) and structural zero analysis.
https://doi.org/10.1371/journal.pone.0335007.s007
(DOCX)
S8 Table. ANCOM-BC2 results of differential abundance of fungal ASVs between soil A and soil B in endosphere samples based on Log Fold Change (LogFC) and structural zero analysis.
https://doi.org/10.1371/journal.pone.0335007.s008
(DOCX)
S9 Table. Total collection of fungal isolates inhibiting P. infestans, in vitro.
https://doi.org/10.1371/journal.pone.0335007.s009
(DOCX)
S10 Table. Collection of 61 identified fungal biocontrol candidates inhibiting P. infestans after eliminating potential plant and human pathogens based on literature search.
https://doi.org/10.1371/journal.pone.0335007.s010
(DOCX)
S11 Table. Collection of 165 unidentified fungal isolates after eliminating those growing at 37˚C and non viable isolates.
https://doi.org/10.1371/journal.pone.0335007.s011
(XLSX)
S12 Table. Collection of 246 initial fungal isolates recovered from tomato plant tissues.
https://doi.org/10.1371/journal.pone.0335007.s012
(XLSX)
S1 Fig. Rarefaction curve of observed species richness after sequencing of ITS rRNA gene from the rhizosphere, endosphere, and phyllosphere samples of tomato plants grown in soil origins A and B.
https://doi.org/10.1371/journal.pone.0335007.s013
(DOCX)
Acknowledgments
The authors thank Anja Götzmann, Julia Herzberg, and Ramona Mittag for technical support and Julian Maier for assistance with the analysis of plant trial data. Christian Fetzer from Landesbetrieb Landwirtschaft Hessen (LLH) established the contact to the growers whose soil was used for the study, which is kindly acknowledged.
References
- 1. Ali MY, Sina AAI, Khandker SS, Neesa L, Tanvir EM, Kabir A, et al. Nutritional composition and bioactive compounds in tomatoes and their impact on human health and disease: a review. Foods. 2020;10(1):45. pmid:33375293
- 2. Panno S, Davino S, Caruso AG, Bertacca S, Crnogorac A, Mandić A, et al. A review of the most common and economically important diseases that undermine the cultivation of tomato crop in the mediterranean basin. Agronomy. 2021;11(11):2188.
- 3. Hashemi M, Tabet D, Sandroni M, Benavent-Celma C, Seematti J, Andersen CB, et al. The hunt for sustainable biocontrol of oomycete plant pathogens, a case study of Phytophthora infestans. Fungal Biology Reviews. 2022;40:53–69.
- 4. Cooke LR, Schepers HTAM, Hermansen A, Bain RA, Bradshaw NJ, Ritchie F, et al. Epidemiology and integrated control of potato late blight in Europe. Potato Res. 2011;54(2):183–222.
- 5. Lehsten V, Wiik L, Hannukkala A, Andreasson E, Chen D, Ou T, et al. Earlier occurrence and increased explanatory power of climate for the first incidence of potato late blight caused by Phytophthora infestans in Fennoscandia. PLoS One. 2017;12(5):e0177580. pmid:28558041
- 6. Fry WE, Birch PRJ, Judelson HS, Grünwald NJ, Danies G, Everts KL, et al. Five reasons to consider phytophthora infestans a reemerging pathogen. Phytopathology. 2015;105(7):966–81. pmid:25760519
- 7. Haverkort AJ, Struik PC, Visser RGF, Jacobsen E. Applied biotechnology to combat late blight in potato caused by phytophthora infestans. Potato Res. 2009;52(3):249–64.
- 8. Kamoun S, Furzer O, Jones JDG, Judelson HS, Ali GS, Dalio RJD, et al. The Top 10 oomycete pathogens in molecular plant pathology. Mol Plant Pathol. 2015;16(4):413–34. pmid:25178392
- 9. Wan N-F, Fu L, Dainese M, Kiær LP, Hu Y-Q, Xin F, et al. Pesticides have negative effects on non-target organisms. Nat Commun. 2025;16(1):1360. pmid:39948065
- 10. Wang Y, Tyler BM, Wang Y. Defense and Counterdefense During Plant-Pathogenic Oomycete Infection. Annu Rev Microbiol. 2019;73:667–96. pmid:31226025
- 11.
European Parliament. Directive 2009/128/EC of the European Parliament and the Council of 21 October 2009 establishing a framework for community action to achieve the sustainable use of pesticides. 2009;71–86.
- 12. Pieterse CMJ, Zamioudis C, Berendsen RL, Weller DM, Van Wees SCM, Bakker PAHM. Induced systemic resistance by beneficial microbes. Annu Rev Phytopathol. 2014;52:347–75. pmid:24906124
- 13. Spadaro D, Droby S. Development of biocontrol products for postharvest diseases of fruit: The importance of elucidating the mechanisms of action of yeast antagonists. Trends Food Sci Technol. 2016;47:39–49.
- 14. Raaijmakers JM, Mazzola M. Diversity and natural functions of antibiotics produced by beneficial and plant pathogenic bacteria. Annu Rev Phytopathol. 2012;50:403–24. pmid:22681451
- 15. Khabbaz SE, Zhang L, Cáceres LA, Sumarah M, Wang A, Abbasi PA. Characterisation of antagonistic Bacillus and Pseudomonas strains for biocontrol potential and suppression of damping-off and root rot diseases. Ann Appl Biol. 2015;166(3): 456–71.
- 16. Chowdappa P, Mohan Kumar SP, Jyothi Lakshmi M, Upreti KK. Growth stimulation and induction of systemic resistance in tomato against early and late blight by Bacillus subtilis OTPB1 or Trichoderma harzianum OTPB3. Biological Control. 2013;65(1):109–17.
- 17. De Silva NI, Brooks S, Lumyong S, Hyde KD. Use of endophytes as biocontrol agents. Fungal Biology Reviews. 2019;33(2):133–48.
- 18.
Rai S, Solanki MK, Solanki AC, Surapathrudu K. Biocontrol potential of Trichoderma spp.: current understandings and future outlooks on molecular techniques. In: Ansari RA, Mahmood I, editors. Plant Health Under Biotic Stress. Microbial Interactions. Singapore: Springer Singapore; 2019. 2. p. 129–160.
- 19. Luo C, He Y, Chen Y. Rhizosphere microbiome regulation: Unlocking the potential for plant growth. Curr Res Microb Sci. 2024;8:100322. pmid:39678067
- 20. Bakker PAHM, Pieterse CMJ, de Jonge R, Berendsen RL. The Soil-Borne Legacy. Cell. 2018;172(6):1178–80. pmid:29522740
- 21. Gush S, Lebre P, Couthino TA, Cowan DA, van der Waals JE. Disentangling shifts in the soil microbiome of potatoes infected with rhizoctonia solani anastomosis group 3-PT in search of potential biocontrol agents. Phytobiomes Journal. 2024;8(2):128–42.
- 22. Raaijmakers JM, Paulitz TC, Steinberg C, Alabouvette C, Moënne-Loccoz Y. The rhizosphere: a playground and battlefield for soilborne pathogens and beneficial microorganisms. Plant Soil. 2008;321(1–2):341–61.
- 23. Jones DL, Hinsinger P. The rhizosphere: complex by design. Plant Soil. 2008;312(1–2):1–6.
- 24. Berendsen RL, Pieterse CMJ, Bakker PAHM. The rhizosphere microbiome and plant health. Trends Plant Sci. 2012;17(8):478–86. pmid:22564542
- 25. Bakker PAHM, Berendsen RL, Doornbos RF, Wintermans PCA, Pieterse CMJ. The rhizosphere revisited: root microbiomics. Front Plant Sci. 2013;4:165. pmid:23755059
- 26. Jayne B, Quigley M. Influence of arbuscular mycorrhiza on growth and reproductive response of plants under water deficit: a meta-analysis. Mycorrhiza. 2014;24(2):109–19. pmid:23917611
- 27. Baum C, El-Tohamy W, Gruda N. Increasing the productivity and product quality of vegetable crops using arbuscular mycorrhizal fungi: A review. Sci Hortic. 2015;187:131–41.
- 28. El Komy MH, Saleh AA, Eranthodi A, Molan YY. Characterization of novel trichoderma asperellum isolates to select effective biocontrol agents against tomato fusarium wilt. Plant Pathol J. 2015;31(1):50–60. pmid:25774110
- 29. Treseder KK, Lennon JT. Fungal traits that drive ecosystem dynamics on land. Microbiol Mol Biol Rev. 2015;79(2):243–62. pmid:25971588
- 30. Brar B, Bala K, Saharan BS, Sadh PK, Duhan JS. Bio-boosting agriculture: Harnessing the potential of fungi-bacteria-plant synergies for crop improvement. Discov Plants. 2024;1(1).
- 31. Kim H-Y, Choi GJ, Lee HB, Lee S-W, Lim HK, Jang KS, et al. Some fungal endophytes from vegetable crops and their anti-oomycete activities against tomato late blight. Lett Appl Microbiol. 2007;44(3):332–7. pmid:17309513
- 32. Köhl J, Postma J, Nicot P, Ruocco M, Blum B. Stepwise screening of microorganisms for commercial use in biological control of plant-pathogenic fungi and bacteria. Biological Control. 2011;57(1):1–12.
- 33. Drenker C, El Mazouar D, Bücker G, Weißhaupt S, Wienke E, Koch E, et al. Characterization of a disease-suppressive isolate of lysobacter enzymogenes with broad antagonistic activity against bacterial, oomycetal and fungal pathogens in different crops. Plants (Basel). 2023;12(3):682. pmid:36771766
- 34. Wieland G, Neumann R, Backhaus H. Variation of microbial communities in soil, rhizosphere, and rhizoplane in response to crop species, soil type, and crop development. Appl Environ Microbiol. 2001;67(12):5849–54. pmid:11722945
- 35.
Semagn K. Leaf tissue sampling and dna extraction protocols. Molecular plant taxonomy: methods and protocols. Totowa, NJ: Humana Press. 2014. p. 53–67.
- 36. White TJ, Bruns T, Lee S, Taylor J. Amplification and direct sequencing of fungal ribosomal rna genes for phylogenetics. PCR Protocols. Elsevier. 1990. p. 315–22.
- 37. Madden T. The BLAST sequence analysis tool. NCBI Handb. 2013;2(5):425–36.
- 38. Linkies A, Jacob S, Zink P, Maschemer M, Maier W, Koch E. Characterization of cultural traits and fungicidal activity of strains belonging to the fungal genus Chaetomium. J Appl Microbiol. 2021;131(1):375–91. pmid:33249672
- 39. Miller AN, Huhndorf SM. Multi-gene phylogenies indicate ascomal wall morphology is a better predictor of phylogenetic relationships than ascospore morphology in the Sordariales (Ascomycota, Fungi). Mol Phylogenet Evol. 2005;35(1):60–75. pmid:15737582
- 40. O’Donnell K, Cigelnik E. Two divergent intragenomic rDNA ITS2 types within a monophyletic lineage of the fungus Fusarium are nonorthologous. Mol Phylogenet Evol. 1997;7(1):103–16. pmid:9007025
- 41. Gardes M, Bruns TD. ITS primers with enhanced specificity for basidiomycetes--application to the identification of mycorrhizae and rusts. Mol Ecol. 1993;2(2):113–8. pmid:8180733
- 42. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011;17(1):10–2.
- 43. Afgan E, Baker D, Batut B, van den Beek M, Bouvier D, Cech M, et al. The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update. Nucleic Acids Res. 2018;46(W1):W537–44. pmid:29790989
- 44.
R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. 2022.
- 45. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13(7):581–3. pmid:27214047
- 46. Abarenkov K, Nilsson RH, Larsson KH, Taylor AFS, May TW, Frøslev TG, et al. The UNITE database for molecular identification and taxonomic communication of fungi and other eukaryotes: sequences, taxa and classifications reconsidered. Nucleic Acids Res. 2024;52(D1): D791–7.
- 47. Fernandez-Gnecco G, Gégu L, Covacevich F, Consolo VF, Bouffaud ML, Buscot F. Alone as effective as together: AMF and Trichoderma inoculation boost maize performance but differentially shape soil and rhizosphere microbiota. J Sustain Agric Environ. 2024;3(1):1–16.
- 48. Wickham H, Averick M, Bryan J, Chang W, McGowan L, François R, et al. Welcome to the Tidyverse. JOSS. 2019;4(43):1686.
- 49.
Wickham H. ggplot2: Elegant graphics for data analysis. 2016.
- 50.
Wilke CO, Wienik BM. ggtext: Improved text rendering support for ‘ggplot2’. 2022.
- 51.
Wickham H, François R, Henry L, Müller K. dplyr: A grammar of data manipulation. R Package. 2023.
- 52.
Müller K, Wickham H. Tibble: Simple data frames.
- 53.
Wickham H. stringr: Simple, consistent wrappers for common string operations. 2019.
- 54.
Elkin LA, Kay M, Higgins JJ, Wobbrock JO. An Aligned Rank Transform Procedure for Multifactor Contrast Tests. In: The 34th Annual ACM Symposium on User Interface Software and Technology, 2021. 754–68. https://doi.org/10.1145/3472749.3474784
- 55.
Oksanen J, Simpson G, Blanchet F, Kindt R, Legendre P, Minchin P. Vegan: community ecology package. 2024.
- 56. McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8(4):e61217. pmid:23630581
- 57. Schloss PD. Sequencing effort in amplicon sequence analyses. mSphere. 2024;0(0).
- 58. Lin H, Peddada SD. Analysis of compositions of microbiomes with bias correction. Nat Commun. 2020;11(1):3514. pmid:32665548
- 59.
VDLUFA. Phosphordüngung nach Bodenuntersuchung und Pflanzenbedarf. Darmstadt: VDLUFA-Verlag. 2018.
- 60. Koch E, Ole Becker J, Berg G, Hauschild R, Jehle J, Köhl J, et al. Biocontrol of plant diseases is not an unsafe technology!. J Plant Dis Prot. 2018;125(2):121–5.
- 61. Pandey AK, Kumar A, Dinesh K, Varshney R, Dutta P. The hunt for beneficial fungi for tomato crop improvement – Advantages and perspectives. Plant Stress. 2022;6:100110.
- 62. Schisler DA, Slininger PJ. Microbial selection strategies that enhance the likelihood of developing commercial biological control products. J Ind Microbiol Biotechnol. 1997;19(3):172–9.
- 63. Philippot L, Raaijmakers JM, Lemanceau P, van der Putten WH. Going back to the roots: the microbial ecology of the rhizosphere. Nat Rev Microbiol. 2013;11(11):789–99. pmid:24056930
- 64. Bhagat S, Shete P, Jain A. Plant health: Feedback effect of root exudates and rhizobiome interactions. Rhizobiome Ecol Manag Appl. 2023;:345–75.
- 65. Vorholt JA. Microbial life in the phyllosphere. Nat Rev Microbiol. 2012;10(12):828–40. pmid:23154261
- 66. Rastogi G, Coaker GL, Leveau JHJ. New insights into the structure and function of phyllosphere microbiota through high-throughput molecular approaches. FEMS Microbiol Lett. 2013;348(1):1–10. pmid:23895412
- 67. Fürnkranz M, Wanek W, Richter A, Abell G, Rasche F, Sessitsch A. Nitrogen fixation by phyllosphere bacteria associated with higher plants and their colonizing epiphytes of a tropical lowland rainforest of Costa Rica. ISME J. 2008;2(5):561–70. pmid:18273066
- 68. Lalaymia I, Naveau F, Arguelles Arias A, Ongena M, Picaud T, Declerck S. Screening and efficacy evaluation of antagonistic fungi against Phytophthora infestans and combination with arbuscular mycorrhizal fungi for biocontrol of late blight in potato. Front Agron. 2022;4:1–20.
- 69.
Pitt JI. Penicillium | Penicillium and Talaromyces: Introduction. Second Edi ed. Oxford: Academic Press. 2014.
- 70. Khan R, Anwar F, Ghazali FM. A comprehensive review of mycotoxins: Toxicology, detection, and effective mitigation approaches. Heliyon. 2024;10(8):e28361. pmid:38628751
- 71. Harman GE, Howell CR, Viterbo A, Chet I, Lorito M. Trichoderma species--opportunistic, avirulent plant symbionts. Nat Rev Microbiol. 2004;2(1):43–56. pmid:15035008
- 72. Adnan M, Islam W, Shabbir A, Khan KA, Ghramh HA, Huang Z, et al. Plant defense against fungal pathogens by antagonistic fungi with Trichoderma in focus. Microb Pathog. 2019;129:7–18. pmid:30710672
- 73.
Funck Jensen D, Dubey M, Jensen B, Karlsson M. Clonostachys rosea to control plant diseases. 2021;:429–72.
- 74. Egel DS, Hoagland L, Davis J, Marchino C, Bloomquist M. Efficacy of organic disease control products on common foliar diseases of tomato in field and greenhouse trials. Crop Protection. 2019;122:90–7.
- 75. Høyer AK, Jørgensen HJL, Amby DB, Jensen B. Endophytic colonization of tomato plants by the biological control agent Clonostachys rosea. Biol Integr Control plant Pathog IOBC-WPRS Bull. 2016;115:101–6.
- 76. European Food Safety Authority. Conclusion on the peer review of the pesticide risk assessment of the active substance Trichoderma gamsii ICC080. EFSA Journal. 2015;13(4).
- 77. Arena M, Auteri D, Barmaz S, Bellisai G, Brancato A, Brocca D, et al. Peer review of the pesticide risk assessment of the active substance Clonostachys rosea strain J1446 (approved in Regulation (EU) No 540/2011 as Gliocladium catenulatum strain J1446). EFSA J. 2017;15(7).
- 78. Manzotti A, Bergna A, Burow M, Jørgensen HJL, Cernava T, Berg G, et al. Insights into the community structure and lifestyle of the fungal root endophytes of tomato by combining amplicon sequencing and isolation approaches with phytohormone profiling. FEMS Microbiol Ecol. 2020;96(5):fiaa052. pmid:32239208
- 79. Ibrahim SRM, Mohamed SGA, Sindi IA, Mohamed GA. Biologically active secondary metabolites and biotechnological applications of species of the family Chaetomiaceae (Sordariales): an updated review from 2016 to 2021. Mycol Progress. 2021;20(5):595–639.
- 80. Kedves O, Kocsubé S, Bata T, Andersson MA, Salo JM, Mikkola R, et al. Chaetomium and Chaetomium-like Species from European Indoor Environments Include Dichotomopilus finlandicus sp. nov. Pathogens. 2021;10(9):1133. pmid:34578165
- 81. Liu C, Kan S, Zhou H, Zhang H, Chang Z. Mycolytic effect and underlying mechanism of Chaetomium subaffine LB-1 on phytopathogenic fungi via morphological, physio-biochemical and transcriptomic analysis. J Phytopathol. 2024;172(1):1–12.
- 82. Stephan D, Schmitt A, Carvalho SM, Seddon B, Koch E. Evaluation of biocontrol preparations and plant extracts for the control of Phytophthora infestans on potato leaves. Eur J Plant Pathol. 2005;112(3):235–46.
- 83. Berg G, Rybakova D, Fischer D, Cernava T, Vergès M-CC, Charles T, et al. Microbiome definition re-visited: old concepts and new challenges. Microbiome. 2020;8(1):103. pmid:32605663
- 84. Gruet C, Muller D, Moënne-Loccoz Y. Significance of the diversification of wheat species for the assembly and functioning of the root-associated microbiome. Front Microbiol. 2022;12:782135. pmid:35058901
- 85. Sahu KP, Patel A, Kumar M, Sheoran N, Mehta S, Reddy B, et al. Integrated metabarcoding and culturomic-based microbiome profiling of rice phyllosphere reveal diverse and functional bacterial communities for blast disease suppression. Front Microbiol. 2021;12:780458. pmid:34917058
- 86.
Watson A, Napier T, Mazwan Muhammad R, Rozana Nik Mohd Masdek N, Zahrah Ponari S, Anak Makup J. Diseases of cucurbit vegetables. 832. 2009;2–3
- 87. Ogórek R, Lejman A, Pusz W, Miłuch A, Miodyńska P. Characteristics and taxonomy of Cladosporium fungi. Mikol Lek. 2012;19(2):80–5.
- 88. Hamayun M, Afzal Khan S, Ahmad N, Tang D-S, Kang S-M, Na C-I, et al. Cladosporium sphaerospermum as a new plant growth-promoting endophyte from the roots of Glycine max (L.) Merr. World J Microbiol Biotechnol. 2009;25(4):627–32.
- 89. Sandoval-Denis M, Sutton DA, Martin-Vicente A, Cano-Lira JF, Wiederhold N, Guarro J, et al. Cladosporium species recovered from clinical samples in the United States. J Clin Microbiol. 2015;53(9):2990–3000. pmid:26179305
- 90. Shakya M, Gottel N, Castro H, Yang ZK, Gunter L, Labbé J, et al. A multifactor analysis of fungal and bacterial community structure in the root microbiome of mature Populus deltoides trees. PLoS One. 2013;8(10):e76382. pmid:24146861
- 91. Pfeiffer S, Mitter B, Oswald A, Schloter-Hai B, Schloter M, Declerck S, et al. Rhizosphere microbiomes of potato cultivated in the High Andes show stable and dynamic core microbiomes with different responses to plant development. FEMS Microbiol Ecol. 2017;93(2):fiw242. pmid:27940644
- 92. Taffner J, Cernava T, Erlacher A, Berg G. Novel insights into plant-associated archaea and their functioning in arugula (Eruca sativa Mill.). J Adv Res. 2019;19:39–48. pmid:31341668
- 93. Lemanceau P, Blouin M, Muller D, Moënne-Loccoz Y. Let the Core Microbiota Be Functional. Trends Plant Sci. 2017;22(7):583–95. pmid:28549621
- 94. Pascale A, Proietti S, Pantelides IS, Stringlis IA. Modulation of the Root Microbiome by Plant Molecules: The Basis for Targeted Disease Suppression and Plant Growth Promotion. Front Plant Sci. 2020;10:1741. pmid:32038698
- 95. Berg G, Köberl M, Rybakova D, Müller H, Grosch R, Smalla K. Plant microbial diversity is suggested as the key to future biocontrol and health trends. FEMS Microbiol Ecol. 2017;93(5):10.1093/femsec/fix050. pmid:28430944
- 96. Lindow SE, Brandl MT. Microbiology of the phyllosphere. Appl Environ Microbiol. 2003;69(4):1875–83. pmid:12676659
- 97. Ramudingana P, Mamphogoro TP, Kamutando CN, Maboko MM, Modika KY, Moloto KW, et al. Antagonistic potential of endophytic fungal isolates of tomato (Solanum lycopersicum L.) fruits against post-harvest disease-causing pathogens of tomatoes: An in vitro investigation. Fungal Biol. 2024;128(4):1847–58. pmid:38876537
- 98. El-Hasan A, Ngatia G, Link TI, Voegele RT. Isolation, Identification, and Biocontrol Potential of Root Fungal Endophytes Associated with Solanaceous Plants against Potato Late Blight (Phytophthora infestans). Plants (Basel). 2022;11(12):1605. pmid:35736756
- 99. Peixoto RS, Voolstra CR, Sweet M, Duarte CM, Carvalho S, Villela H, et al. Harnessing the microbiome to prevent global biodiversity loss. Nat Microbiol. 2022;7(11):1726–35. pmid:35864220
- 100. Larousse M, Rancurel C, Syska C, Palero F, Etienne C, Industri B, et al. Tomato root microbiota and Phytophthora parasitica-associated disease. Microbiome. 2017;5(1):56. pmid:28511691
- 101. Berg G, Smalla K. Plant species and soil type cooperatively shape the structure and function of microbial communities in the rhizosphere. FEMS Microbiol Ecol. 2009;68(1):1–13. pmid:19243436
- 102. Docherty KM, Borton HM, Espinosa N, Gebhardt M, Gil-Loaiza J, Gutknecht JLM, et al. Key Edaphic properties largely explain temporal and geographic variation in soil microbial communities across four biomes. PLoS One. 2015;10(11):e0135352. pmid:26536666
- 103. Zhang T, Wang N-F, Liu H-Y, Zhang Y-Q, Yu L-Y. Soil pH is a key determinant of soil fungal community composition in the Ny-Ålesund region, Svalbard (High Arctic). Front Microbiol. 2016;7:227. pmid:26955371
- 104. Bardgett RD, van der Putten WH. Belowground biodiversity and ecosystem functioning. Nature. 2014;515(7528):505–11. pmid:25428498