Plant chemical variation mediates soil bacterial community composition

Abstract

An important challenge in the study of ecosystem function is resolving how plant antiherbivore chemical defence expression may influence plant-associated microbes, and nutrient release. We report on a factorial experiment that explores a mechanism underlying this interplay using individuals of the perennial plant Tansy that vary genotypically in the chemical content of their antiherbivore defenses (chemotypes). We assessed to what extent soil and its associated microbial community versus chemotype-specific litter determined the composition of the soil microbial community. Microbial diversity profiles revealed sporadic effects of chemotype litter and soil combinations. Soil source and litter type both explained the microbial communities decomposing the litter with soil source having a more important effect. Some microbial taxa are related to particular chemotypes, and thus intra-specific chemical variation of a single plant chemotype can shape the litter microbial community. But we found that ultimately the effect of fresh litter inputs from a chemotype appeared to act secondary as a filter on the composition of the microbial community, with the primary factor being the existing microbial community in the soil.

Figure 1

Tansy field (A), Tansy plant (B), U. tanaceti aphids on tansy leaves (C) (Photos made by Adalbert Balog), litter decomposition and soil bacterial community assay design (D). Numbers represents sample orders, which were used consequently in labelling samples for genetic analyses. The first four treatments representing the main chemotype soils (I, II, III, IV), the control with no tansy (V) soil were used in boxes, within each, litter additions from the different chemotypes (I-IV) or non-Tansy plants (V) were placed. Marks as S1…S25 represents particular chemotype samples in particular chemotype soil (i.e. III.S3 means Carvone litter in Camphor soil, versus III S8 Carvone litter in Borneol soil).

Figure 2

The 13 most abundant bacterial phyla (A) and genera (B) under different tansy chemotypes with own soil and with other chemotypes soil and control (no tansy). Relative abundance does not sum to 1 to account for reads that did not belong to the 13 most abundant taxa.

Figure 3

Diversity orders according to alpha diversity profiles of bacterial communities under different tansy chemotypes with own soil and with other chemotypes soil and control (no tansy).

Figure 4

Soil and bacterial community interaction at the Phylum (A) and at the Genus (B) level. Non-Metric Multidimensional Scaling (NMDS) using microbial sequencing data were applied to reveal relational patterns among bacterial phyla and genera and tansy chemotype. The large points show where soil and litter came from the same chemotype or control site.