Microbial Interactions in the Phyllosphere Increase Plant Performance under Herbivore Biotic Stress

Abstract

The phyllosphere supports a tremendous diversity of microbes and other organisms. However, little is known about the colonization and survival of pathogenic and beneficial bacteria alone or together in the phyllosphere across the whole plant life-cycle under herbivory, which hinders our ability to understand the role of phyllosphere bacteria on plant performance. We addressed these questions in experiments using four genetically and biogeographically diverse accessions of Arabidopsis thaliana, three ecologically important bacterial strains (Pseudomonas syringae DC3000, Xanthomonas campestris, both pathogens, and Bacillus cereus, plant beneficial) under common garden conditions that included fungus gnats (Bradysia spp.). Plants supported greater abundance of B. cereus over either pathogenic strain in the phyllosphere under such greenhouse conditions. However, the Arabidopsis accessions performed much better (i.e., early flowering, biomass, siliques, and seeds per plant) in the presence of pathogenic bacteria rather than in the presence of the plant beneficial B. cereus. As a group, the plants inoculated with any of the three bacteria (Pst DC3000, Xanthomonas, or Bacillus) all had a higher fitness than uninoculated controls under these conditions. These results suggest that the plants grown under the pressure of different natural enemies, such as pathogens and an herbivore together perform relatively better, probably because natural enemies induce host defense against each other. However, in general, a positive impact of Bacillus on plant performance under herbivory may be due to its plant-beneficial properties. In contrast, bacterial species in the mixture (all three together) performed poorer than as monocultures in their total abundance and host plant growth promotion, possibly due to negative interspecific interactions among the bacteria. However, bacterial species richness linearly promoted seed production in the host plants under these conditions, suggesting that natural enemies diversity may be beneficial from the host perspective. Collectively, these results highlight the importance of bacterial community composition on plant performance and bacterial abundance in the phyllosphere.

Keywords: antagonistic interactions; bacterial species richness; beneficial and pathogenic bacteria; biotic stress; herbivory; phyllosphere; plant performance; plant–microbe–insect interactions.

FIGURE 1

Abundance of bacterial monocultures in the phyllosphere of Arabidopsis accessions. The performance of bacterial monocultures in the phyllosphere of all Arabidopsis accessions (A). The comparison between the performance of beneficial (Bacillus) and pathogenic (Pst DC3000 and Xanthomonas together) bacteria in the phyllosphere of all Arabidopsis accessions (B). The monoculture treatments are the average of three replicates, whereas the panel (B) shows an average of all replicates in the both beneficial and pathogenic treatments. The significant differences were determined by ANOVA followed by Tukey’s test. Error bars represent means ± 1SE. Lack of shared letters above the bars indicate significant differences.

FIGURE 2

Impact of individual bacterial species on the plant performance. Effect of bacterial monoculture treatments on (A) days to flowering (DF), (B) dry biomass per plant, (C) siliques per plant (SPP), and (D) seeds per silique (SPS) production in all Arabidopsis accessions (NFA-8, Ba1-2, Kelst-4, and Tu-0). Sample size varied from 6 to 15 plants (e.g., Kover and Schaal, 2002). Error bars represent means ± 1SE. The significant differences were determined by ANOVA followed by Tukey’s, and Fisher’s tests. Lack of shared letters above the bars indicate significant difference.

FIGURE 3

Impact of bacterial species richness on the bacterial abundance, and the relative size of diversity effects in bacterial communities across plant accessions. Effects of bacterial species richness on the bacterial abundance in the phyllosphere of all Arabidopsis accessions (A). All monoculture and mixture treatments were in triplicate. Significant differences were determined by ANOVA followed by Tukey’s test. Error bars represent means ± 1SE. The tripartite equations were applied to calculate the (B) transgressive yield, (C,D) net biodiversity effect (NBE), complementarity effect (CE), and selection effect (SE) for each genotype, and (E) averaged diversity effects across all accessions.

FIGURE 4

A comparison of the performance of bacterial species in the monocultures and mixtures across all plant accessions. The comparative performance of all bacterial species in the monoculture and mixture across all Arabidopsis accessions. All monoculture and mixture treatments were in triplicate. Significant differences were determined by ANOVA followed by Fisher’s test. Error bars represent means ± 1SE.

FIGURE 5

Impact of bacterial species richness on plant performance. Effect of bacterial species richness (monoculture vs mixture) on (A) DF, (B) dry biomass per plant, and (C) SPP production in all Arabidopsis accessions. Sample size varied from 6 to 15 plants (e.g., Kover and Schaal, 2002). Error bars represent means ± 1SE. The significant differences were determined by ANOVA followed by Fisher’s tests. Lack of shared letters above the bars indicate significant difference.

FIGURE 6

Impact of bacterial species richness on the SPS production across all plant accessions. Observed seed per silique production in all Arabidopsis accessions as a function of bacterial species richness across all treatments. Each point in the figure corresponds to the average of various replicates (mostly > 3). The points mentioning to respective treatments at equal richness level are slightly offset horizontally for clarity. The ANOVA with a linear fitting on the means was performed to determine the effect of bacterial species richness on SPS.