Non-host larvae negatively impact persistence of the entomopathogen Beauveria bassiana in soil

A better understanding of the ecology of the insect pathogenic fungus, Beauveria bassiana, in soil is needed to identify reasons behind the variable efficacy often seen after field application. A transformed strain of a candidate commercial strain of B. bassiana (F418 gfp tr3), expressing the green fluorescent protein and the hygromycin B resistance gene, was used to assess the effects of the larvae of a host insect, Tenebrio molitor L. (Coleoptera: Tenebrionidae), a non-host, Costelytra zealandica (Coleoptera: Scarabaeidae) and the absence of larvae on the persistence of F418 gfp tr3 in pasteurised and non-sterile soil over 4 months.In the presence of a T. molitor larvae, F418 gfp tr3 populations increased significantly in pasteurised and non-sterile soil; however, populations increased less in non-sterile soil than in pasteurised soil. Lower populations of F418 gfp tr3 were recovered in pasteurised soil in the presence of C. zealandica larvae than in pasteurised soil without larvae. No difference was observed between F418 gfp tr3 populations in non-sterile soil with a non-host larvae or without larvae.Accompanying studies showed that F418 gfp tr3 conidia germinated and produced appressoria on live and excised cuticle of non-host (C. zealandica) larvae but infection did not occur, leading to a net loss of viable conidia in the soil. Conidia administrated orally to C. zealandica larvae were viable on recovery from faecal samples, suggesting that ingestion of the fungus by the larvae had little impact on the viable fungal population.Soil bacterial and fungal community patterns were analysed using Single-Strand Conformation Polymorphism (SSCP) and showed a correlation between changes in F418 gfp tr3 persistence in pasteurised and non-sterile soil and changes in soil communities in the presence of a host insect, non-host insect or in the absence of insect.In pasteurised soil, non-specific germination of F418 gfp tr3 conidia on the non-host larval cuticle and the presence of antagonistic bacteria introduced with the field-collected larvae are most likely responsible for the differences observed. The more complex microbial community structures in non-sterile soil could lead to fungistasis, preventing potentially antagonistic bacteria degrading conidia or inhibiting attachment and germination on the non-host larval cuticle, resulting in the observed lack of difference between non-host and no larval treatments.