The capacity to suppress wheat crown rot disease occurs naturally in contrasted farm soils

Context This repositery accompanies the submitted manuscript "The capacity to suppress wheat crown rot disease occurs naturally in contrasted farm soils". Authors: Alix Catrya, Simon De Dondera, Richard Blancb, Michaël Denefleb, Romain Bouquetb, Géraldine Philippeb, Stéphane Bernardb, Thierry Langinb, Karim Gharred-Noelc, Lionel Lebretonc, Christophe Mougelc, Stéphane Cordeaud, Marie-Charlotte Boppd, Chrystel Gibelin-Vialae, Oriane Bessete, Fabienne Vailleaue, Fabrice Rouxe, Yvan Moënne-Loccoza,f, Daniel Mullera aUniversité Claude Bernard Lyon 1, CNRS, INRAE, VetAgro Sup, UMR5557 Ecologie Microbienne, 43 bd du 11 novembre 1918, F-69622 Villeurbanne, France bINRAE/UBP, UMR1095 Genetics, Diversity and Ecophysiology of Cereals (GDEC), F-63000 Clermont-Ferrand, France cINRAE, Agrocampus Ouest, Université de Rennes, IGEPP, F-35650 Le Rheu, France dUniversité de Bourgogne Europe, Institut Agro Dijon, INRAE, UMR Agroécologie, F-21000 Dijon, France. eLIPME, Université de Toulouse, INRAE, CNRS, F-31320 Castanet-Tolosan, France fInstitut Universitaire de France (IUF), F-75005 Paris, France Experiments Screening of 49 soils for their suppressiveness to Fusarium crown rot of wheat in greenhouse experiments (inoculated or not with the pathogen, n = 10 pots per soil & treatment, with 4 plants per pot). The soils were sampled in the begginning of 2023 in three regions in France (South-West, North-East and North-West) and another soil near Clermont-Ferrand (France). Figures Fig. 1. Mean disease score for F. graminearum-mediated crown rot symptoms in wheat plants grown in 49 different soils, inoculated or non-inoculated with F. graminearum Fg1. For each region, bars are in ascending order of the average symptom score in inoculated conditions. Stars indicate significant differences between treatments for the same soil (100,000 permutations, FDR correction for multiple testing), with * (P < 0.05), ** (P < 0.01) or *** (P < 0.001) and NS, not significant. Significant differences between control and inoculated samples were observed for soils AF152 (P = 0.025), AF154 (P = 0.022), AF162 (P = 5.2 × 10-4), AF165 (P = 0.014), AF166 (P = 0.025), AF167 (P < 1 × 10-4), AF172 (P = 0.0030), AF184 (P = 5.2 × 10-4), AF188 (P = 0.022), AF190 (P = 0.025), AF197 (P = 0.0022), AF198 (P = 0.026), AF199 (P < 1 × 10-4) and for sand for S1 (P = 5.2 × 10-4) and S2 (P < 1 × 10-4). A 46% change in mean symptom severity associated with a fixed effect (soil or inoculation) could be detected, as indicated by a Tweedie GLM model, with a minimal detectable effect size of 0.376 (log scale). Fig. 2. PCA (A) and PLS-DA (B) comparison of the 49 soils based on chemical properties. Organic carbon content was lower overall in suppressive soils than in conducive soils (Wilcoxon rank test, P = 0.034, effect size = 0.40, CI95% = [-9.9, -0.6]) and is highlighted using bold and an asterisk, whereas differences were not significant for the other chemical parameters. When statistics were done with the 28 cambisols or the 11 luvisols, the only difference found was that, in luvisols only, Mg2+ content was significantly lower in suppressive soils than in intermediary or conducive soils combined (Wilcoxon test, P = 0.0467, effect size = 0.63, CI95% = [-5.6, -0.09]). Fig. S1. Geographic positions of the 49 fields where soil was sampled, coloured according to their conducive, intermediate, or suppressive status regarding crown rot caused by F. graminearum. Coordinates of the field plots were retrieved and used to position them on a map of France with R packages ‘rnaturalearthdata’, ‘rnaturalearth’, ‘sf’, and ‘ggspatial’, and the boundaries of France’s counties (départements) were downloaded (in shapefile format) from the data.gouv.fr repository (https://www.data.gouv.fr/fr/datasets/r/eb36371a-761d-44a8-93ec-3d728bec17ce). Soils AF152, AF162, AF163, AF164, AF165, AF166, AF168, AF169 (North-East), AF170, AF171, AF172, AF173, AF174, AF176, AF177 (South-West), AF184, AF187, AF189, AF190, AF195, AF196, AF197, AF198, AF199, AF200 (North-West) were collected in January 2023, and soils AF153, AF154, AF155, AF156, AF157, AF158, AF159, AF160, AF161 (North-East), AF175, AF178, AF179, AF180, AF181, AF182, AF183 (South-West), AF186, AF187, AF188, AF190, AF191, AF192, AF193, AF194 (North-West), AF202 in February 2023. Fig. S2. Disease symptom scale for Fusarium crown rot. Crown rot severity is indicated by a score from 0 to 5, 0 being the absence of symptoms and 5 the most severe disease symptoms. The parts of the stem that are surrounded by a white line correspond to necrosis spots and plaques. Fig. S3. Crown rot severity depending on the region of origin of the soil (Clermont-Ferrand soil not included). Every dot corresponds to a pot, and the score represents the average disease score of plants in the pot. Dots and bars are coloured according to treatment, i.e. whether the plants were inoculated or not with F. graminearum. Tables Table S1. Soil type and chemical properties of the 49 soils studied. [Chemical analysis of soils was carried out using standard procedures for soil texture (pipette method ; NF X31-107), pH (in water), contents in limestone (Bernard calcimeter ; NFX 31-105), organic carbon (dry combustion ; ISO 10694), total N (dry combustion ; ISO 13878), available P (Dyer method for acidic and neutral soils, Olsen method for alkaline soils ; NFX 31-160), Zn, Cu and Mn (ammonium acetate and EDTA, NFX 31-120), and cation exchange capacity (CEC) and levels in exchangeable cations Ca2+, Mg2+, K+ and Na+ (Metson and ammonium acetate ; NFX 31-130). CEC sat, percent saturation of the CEC.] Table S2 Chemical composition of the nutrient solution used in the plant assays. pH was 6.28 and Electrical conductivity 1.33 mS. Scripts soil_screening_stats.Rmd. Identify suppressive, conducive and intermediate soils and investigate links to chemical properties and farming practices of the soils. Permutation tests are used to detect significant differences in symptoms between treatments. PCA and PLS-DA (supervised by the suppressive or conducive status of the soil) are performed with chemical properties/those linked to management of soils as variables. Wilcoxon tests are used to identify significant differences between chemical/soil management variables in suppressive vs conducive soils, and Spearman correlation tests to identify significant correlation between symtom level and soil management variables. soil_screening_stats_GLMM.Rmd. Verifies the suppressive status of soils using another method, i.e. a Tweedie Generalized Linear Mixed Model (Dunn and Smyth 2005) using the pot as a random effect. carteDeepImpact.R. Generate the maps for supplementary figure 1. Data locations_field_plots_approx.xlsx. Lattitude and longitude for plots where the soils were sampled in 2023 – these values are approximated for confidentiality purposes. plant_symptoms_data.xlsx. Per plant symptom grades. sols_notation_pots_DEEP_IMPACT_criblage.csv. Sum of symptoms grades per pot (for all 4 plants). ta_soils_y2_s1_261125_noLoc.csv. Chemical properties of soils with metadata (longitude and latitude removed for confidentiality purposes). Practices_Ta_Y2_ms_criblage_corrected.csv. Variables related to farming/soil management practices.