Fig 1.
Attraction of S. invicta to soil from favored nesting areas and to soil volatiles.
(A) Choices of workers and queens in regard to soil from favored or non-favored nesting areas in olfactometer bioassays (±SE, n = 5 biological replicates). Control: empty arm. (B) Typical gas chromatographic traces obtained from the soil of favored and non-favored nesting areas. (C) Content of geosmin and 2-methylisoborneol (±SE, n = 6 biological replicates, independent sample t test). (D) GC−EAD profile of geosmin and 2-methylisoborneol on worker and queen antenna. Upper trace: GC−FID. Lower trace: EAG (unit: mV). (E) Attraction of S. invicta to geosmin and 2-methylisoborneol in a Y-tube olfactometer (n ≥ 30 biological replicates, chi-square test). Control: mineral oil. Different letters above bars indicate significant differences according to Kendall’s W test at the 0.05 level. Asterisks indicate significant differences (*P < 0.05, **P < 0.01, *** P < 0.001).
Fig 2.
Actinobacterial diversity in soil from favored nesting and non-favored nesting sites.
(A) Taxonomic groups responsible for the differences in microbial diversity between favored and non-favored nesting soil based on LEfSe analysis. (B) Relative abundance of Streptomycetaceae and Nocardiaceae in soil from favored vs. non-favored nesting areas (±SE, n = 6 biological replicates, independent sample t test). (C) Absolute abundance of Streptomyces and Nocardiopsis in soil achieved by qPCR (±SE, n = 6 biological replicates, independent sample t test). Asterisks indicate significant differences (*P < 0.05, ***P < 0.001).
Fig 3.
Volatiles of actinobacterial strains attract S. invicta ants.
(A) Molecular phylogenetic analysis of actinobacterial strains isolated from soil from favored nesting sites and identified based on their 16S rRNA gene sequences. Tree was constructed with a maximum likelihood estimation method. Branch support is indicated as bootstrap values (500 replicates). Red triangles indicate the isolated bacteria. (B) Typical gas chromatographic traces of the headspace of isolated bacterial cultures showing the identified volatiles. (C) Attraction of S. invicta in a Y-tube olfactometer to the isolated bacteria when added to soil from non-favored nesting areas (n ≥ 30 biological replicates, chi-square test). Control: soil from non-favored nesting areas. Asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001).
Fig 4.
Diversity of fungi is greater in soil from areas not favored by nesting S. invicta queens.
(A) Relative abundance of fungi that are pathogenic to animals (±SE, n = 6 biological replicates, independent-sample t tests). (B) and (C) Relative abundance of Clavicipitaceae and Cordycipitaceae (±SE, n = 6 biological replicates, independent-sample t tests). Asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001).
Fig 5.
Entomopathogenic fungi from soil not favored by S. invicta decreases the survival of ant queens.
(A) Kaplan-Meier plot of queen survival after they were forced to build nests in soil from nesting and non-nesting sites (log-rank test). (B) Molecular phylogenetic analysis of fungal strains isolated from soil from non-favored nesting areas based on ITS gene sequences. Tree was constructed with a maximum likelihood estimation method. Branch support is indicated as bootstrap values (500 replicates). Red triangles indicate the isolated fungi. (C) Kaplan-Meier plot of S. invicta queen survival after infection with the isolated fungi (log-rank test). (D) Inhibition effects of the isolated actinobacteria (Fig 3) on the entomopathogenic fungi. On each plate, the fungi were placed in the middle, and the actinobacteria were placed on the four corners of the fungal plug. (E) Kaplan-Meier plot of queen survival after adding the isolated strains of actinobacteria to soil from non-favored nesting sites (log-rank test). Asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001).
Fig 6.
Data from more wild sites confirmed the association between S. invicta queen nesting preference and actinobacteria.
(A) Choices of queens in regard to soil in olfactometer bioassays (±SE, n = 6 biological replicates). Control: empty arm. (B) Content of geosmin and 2-methylisoborneol (±SE, n = 6 biological replicates, independent sample t test). (C) Absolute abundance of Streptomyces and Nocardiopsis in soil achieved by qPCR (±SE, n = 6 biological replicates, independent sample t test). Different letters above bars indicate significant differences according to Kendall’s W test at the 0.05 level. Asterisks indicate significant differences (*P < 0.05, ***P < 0.001).