Thèse Etude de la Diversité Génétique Microbienne pour Améliorer la Conception des Communautés Microbiennes au Service de la Santé du Blé et des Brassicacées H/F - Doctorat.Gouv.Fr

Établissement : Université de Toulouse École doctorale : SEVAB - Sciences Ecologiques, Vétérinaires, Agronomiques et Bioingenieries Laboratoire de recherche : LIPME - Laboratoire des Interactions Plantes-Microbes-Environnement Direction de la thèse : Claudia BARTOLI-KAUTSKY ORCID 0000000314612983 Début de la thèse : 2026-10-01 Date limite de candidature : 2026-06-01T23:59:59 La compréhension des association bénéfiques entre plantes cultivées et microorganismes bénéfiques est indispensable pour identifier les communautés microbiennes les plus efficaces recrutées par la plante et aptes à stimuler les réponses de défense et/ou la croissance de la plante. Dans le contexte de la révolution verte, il est particulièrement pertinent de développer des formulations à base de microorganismes afin de remplacer les pesticides chimiques, aux effets délétères sur les habitats. Bien que plusieurs agents biologiques soient déjà commercialisés, les espèces microbiennes reconnues comme bénéfiques pour les cultures ont été majoritairement développées sans prendre en compte la variation génétique de l'hôte ni celle des microorganismes. De plus, la plupart des biopesticides reposent sur un nombre limité d'espèces microbiennes, avec des effets restreints sur la plante. Pour rendre les biopesticides plus efficaces et durables, il est nécessaire de : i) sélectionner des cultivars portant des loci interagissant positivement avec les microorganismes inoculés ; et ii) considérer des mélanges microbiens complexes produisant une diversité de molécules bénéfiques pour la plante. Grâce à des travaux précédemment réalisés dans l'équipe, nous avons développé une approche permettant d'assembler des mélanges microbiens, ce qui a déjà conduit à l'identification de consortiums présentant des propriétés de biocontrôle supérieures chez la plante par rapport à l'utilisation de souches uniques. Sur la base de ces résultats, ce projet de thèse vise à : i) tester la variabilité intraspécifique de la réponse de souches appartenant à un ensemble d'espèces microbiennes connues pour être bénéfiques en interaction avec les plantes (Trichoderma, Mortierella et Bacillus) ; ii) sélectionner les souches microbiennes les plus efficaces en termes d'activités de biocontrôle et de biostimulation ; iii) combiner plusieurs souches bénéfiques afin de développer des bioproduits efficaces aux propriétés accrues ; et iv) tester ces mélanges sur un ensemble de cultivars de blé en réponse à Fusarium graminearum et sur des populations de Brassica rapa en réponse au champignon pathogène tellurique Rhizoctonia solani. Un des objectif ultime de ce projet de thèse vise à identifier une ou des combinaisons optimales entre microorganismes et plantes afin d'améliorer la croissance de la plante et/ou la résistance à des agents pathogènes fongiques. About three million tonnes of pesticides are used across the globe each year to protect crops with negative consequences on ecosystems and human health. Yet little is known about where or in which environment these chemicals end up after their application on crops. Global consumer trends and markets are driving the need for socially acceptable and environmentally sustainable cropping practices. The unintended environmental impacts of chemical pesticides such as effects on non-target organisms and contamination of soil and waterways is forcing government regulators to phase them out. In addition, climate change and concerns on food security require more resilient cropping systems. Based on their environmental benign attributes, beneficial microbes are an increasingly accepted path forward in the replacement of agrochemicals. Europe is at the forefront of research reducing the use of undesirable agrochemicals, with programs such as the European Union Green Deal. The current commercialized biopesticides are based on a very limited number of microbial species. For example, 90% of microbial biopesticides have been derived from the bacterium Bacillus thuringiensis. Likely for agrochemicals, the formulation of biopesticides does not consider the genetic variability of crops. As a specific molecule or microorganism can have divergent effects on genetically differentiated crops, the pharmaceutical crop industry needs to design specific microbial/crop combinations that take into consideration the genetic architecture of the host. Indeed, a lower efficacy and higher cost are the main disadvantages of many biopesticides compared to chemical pesticides. To make the reduction of chemicals in agriculture possible we have to provide effective and problem-specific solutions. Likewise, the evaluation about the production and persistence of secondary metabolites produced in situ by microbes can be an obstacle for the homologation of microbial formulations. Therefore, clear safety assessments need to be design. The PhD program is defined on the attempt to reduce pesticides and propose solutions based on microbial formulation and the selection of the best host genotype selecting for beneficial microbes. The objectives of the PhD program can be divided into three Tasks:
Task 1. Characterizing intraspecies variability of two bacterial species (Bacillus simplex and Bacillus thuringiensis) and two fungal species (Trichoderma sp. and Mortierella alpina) already characterized for their biocontrol properties on Brassica plants from previous results obtained in the team. Most of the biopesticides on the market are formulated from a limited number of bacterial species and strains. The intraspecies variability among strains is rarely tested. Thanks to bacterial and fungal collections available in the team, we already characterized a large diversity of strains belonging to B. simplex, B. thuringiensis, and Trichoderma sp. We also tested few strains from Trichoderma sp. on Brassica rapa fast cycling plants and found that they are efficient to protect plants against the soil pathogen Rhizoctonia solani. The first task of this PhD project will be to conduct a detailed characterization of these collections. For this, the PhD student will test, in greenhouse conditions, the phenotypic variability of 50 B. simplex, 12 B. thuringiensis, 63 Trichoderma sp. and 50 Mortierella strains for their ability to protect wheat plants against F. graminearum and B. rapa populations against R. solani. In parallel, strains will be also tested for their variability to improve host's growth in green-house conditions.
Task 2. Selecting and designing microbial communities following this first screening in Task1. First, to select for the most effective bacterial and fungal strains to develop microbial community assemblages, statistical analyses, such as linear models, will be applied to identify differences among the strains. Then, to microbial communities will be designed both at the intra-species level, but also at the inter-species level, mixing the most effective fungal and bacterial isolates selected in Task 1. In this aim, to determine the most synergistic associations and in the same time avoid competitive mixture the PhD candidate will will characterize the strains selected in Task 1 for: i) their ability to degrade carbon sources through OmniLogTM ecoplates (for bacteria) and FF plates (for fungi) ; ii) their growth kinetic on Nutrient Agar by using the CLARIOstar automate allowing testing 96 microbe per kinetic ; iii) their ability to degrade IAA (Indole-3-acetic acid) and SA (salicylic acid), two plant hormones related to plant growth and resistance; iv) and their metabolism by whole genome sequencing and metabolic pathways reconstruction based on annotated genomes. The PhD student will integrate all these data to then selected bacterial and fungal strains compatible to establish a consortium. For this, we will notably favor strains that: i) do not compete for carbon sources as they have a divergent OmniLog profile, and ii) harbor properties to degrade plant hormones. For this, the PhD student will use multivariate sparse analysis. In the frame of the PhD project, the student will select a first set of three microbial communities to evaluate their abilities to reduce disease on both wheat and turnip and on their potential to improve the crop growth.
Task 3. Testing three selected microbial communities for their variability on a panel of wheat and turnip varieties. In the last part of the PhD program, the three microbial communities selected at the end of Task 2 will be used to evaluate both their biocontrol and biostimulation effects on a large panel of wheat and turnip varieties. The objective is to better understand the host variability in response to these communities and to identify find the best genotype/microbial community combination. Heritability analyses will be performed to study the host genotype effect on the protection and growth stimulation due to the communities. For the wheat varieties, SNPs matrices are already available and, depending on how much time we have left at the end of the thesis, the PhD student will have the opportunity to perform GWA studies to start the characterization of wheat loci associated with one microbial community.
Task 1. Characterizing intraspecies variability of two bacterial species. The inoculation of the bacterial solution will be done on gamma-ray sterilized soils by using bacterial solution at the stationary phase. The fungal isolates will be inoculated with a toothpick. Both inoculation methods are already in place in the team. For F. graminearum, disease score will be recorded after 4 weeks post incubation by using a rate from 0 to 5 and by measuring the stem necrosis. For R. solani, the disease score (from 0 to 4) will be recorded one week after inoculation. The biostimulation activity will be estimated 2 months after microbial inoculations on both arial and root compartments. Linear-mixed model will be used to test the significance of the fixed effects (e.g. microbial strains) to random effects (replicates, blocks etc.). This first screening will be performed on three genotypes of each species known to have a different gradient of resistance to the pathogens.
Task 2. Designing microbial communities. The microbial collections will be phenotyped for several traits related to competitivity and production of antimicrobial compounds and plant phytohormones. The genomes of all strains will be also sequenced to identify microbial pathways related to biocontrol. Genomic data, phenotypic traits and the data obtained from the green-house characterization will be integrated in a multivariate model and a LASSO model to identify to best microbial communities that reduce disease and improve growth.
Task 3. Testing three selected microbial communities for their variability on a panel of wheat and turnip varieties. The wheat genotypes will be selected from a core collection available at the LIPME laboratory and the CREABio association, the B. rapa populations will be selected on the INRAE CRB located in Brittany. Genotypes of both hosts will be selected to optimize the genetic diversity and for this the PhD student will use genotypic data already produced in the past from other INRAE research teams. The microbial communities will be tested in green-house conditions in a split-plot design suitable for GWA studies. Best-Linear-Prediction means will be estimated and used to estimate the heritability of the response of the microbial communities on the host varieties.

Microbiologie, phytopathologie, statistiques, biologie moléculaire, analyse des génomes fongiques et bactériens