G. Vannacci, C. Cristani, M. Forti, G. Kontoudakis , P. Gambogi
doi: 10.4454/jpp.v81i1.1045
Three different agar media were compared on 13 commercial seed samples of sweet basil to detect Fusarium oxysporum. Komada medium shows a restricted growth of many fungi but only F. oxysporum and species within the Liseola section grow slightly faster and sporulate. Most F. oxysporum colonies were recognizable under the stereomicroscope after training. Komada medium was, therefore, chosen for testing. Out of 21 commercial samples, suspected of harbouring the pathogen, 14 resulted infected and 8 internally infected. Five of these latter gave rise, after sowing, to infected plants demonstrating the seed to plant transmission of the inoculum in commercial seeds. Sowing healthy seeds in soil where infected seed lots had previously been sown gave rise to infected plants supporting the role of infected seeds in soil contamination. Isolates of F. oxysporum from basil seeds were tested for pathogenicity. Most isolates tested (88.2%) belong to the f.sp. basilici but some were saprotrophs, in a few cases they represented the only Fusaria present on seeds. A seed health test based on the agar plate method can lead to the refusal of healthy seed lots actually contaminated by saprotrophic F. oxysporum. Artificial inoculations or VCG determination can furnish a correct identification of the pathogen but they are quite time consuming; a molecular identification of the forma specialis is therefore required.