Skip to main content
. 2023 Jul 24;13:11981. doi: 10.1038/s41598-023-38669-7

Figure 3.

Figure 3

(A) Variable importance obtained through random forest analysis for each bacterial feature in promoting larval growth at day 4 and 7 after bacterial association. (B) Random Forest variable importance results as obtained from ranger function in R (ranger package). Variable importance has been scaled and transformed in percentage to describe percentage contribution of variable to the larval size. (C–F) Larval longitudinal length measured 7 days after inoculation with: (B) live bacterial strains (experimental conditions: − Ery [standard mono-association], + Ery [no daily addition], + Ery [daily addition]), dead bacterial strains (− Ery [daily addition of HT bacteria], + Ery [daily addition of HT bacteria]), and PBS (germ-free GF conditions); (C) the four L. plantarum strains, each including all experimental conditions tested (-Ery [standard mono-association], + Ery [no daily addition], + Ery [daily addition], − Ery [daily addition of HT bacteria], + Ery [daily addition of HT bacteria]); (D) increasing bacterial biomass (growing, experimental conditions: − Ery [standard mono-association], + Ery [daily addition], − Ery [daily addition of HT bacteria], + Ery [daily addition of HT bacteria]) and stable bacterial biomass (+ Ery [no daily addition], − Ery [single inoculation of HT bacteria]); (E) live or dead bacterial strains on LPD with (+ Erythromycin) and without erythromycin (− Erythromycin). Asterisks illustrate statistically significant difference on pairwise intra-strain comparisons (****p < 0.0001, ns: not significant). Center values in the graph represent means and error bars represent SD.