Revealing Cues for Fungal Interplay in the Plant-Air Interface in Vineyards

Abstract

Plant-associated microorganisms play a crucial role in plant health and productivity. Belowground microbial diversity is widely reported as a major factor in determining the composition of the plant microbiome. In contrast, much less is known about the role of the atmosphere in relation to the plant microbiome. The current study examined the hypothesis that the atmospheric microbiome influences the composition of fungal communities of the aboveground organs (flowers, fruit, and leaves) of table grape and vice versa. The atmosphere surrounding grape plantings exhibited a significantly higher level of fungal diversity relative to the nearby plant organs and shared a higher number of phylotypes (5,536 OTUs, 40.3%) with the plant than between organs of the same plant. Using a Bayesian source tracking approach, plant organs were determined to be the major source of the atmospheric fungal community (92%). In contrast, airborne microbiota had only a minor contribution to the grape microbiome, representing the source of 15, 4, and 35% of the fungal communities of leaves, flowers, and fruits, respectively. Moreover, data indicate that plant organs and the surrounding atmosphere shared a fraction of each other's fungal communities, and this shared pool of fungal taxa serves as a two-way reservoir of microorganisms. Microbial association analysis highlighted more positive than negative interactions between fungal phylotypes. Positive interactions were more common within the same environment, while negative interactions appeared to occur more frequently between different environments, i.e., atmosphere, leaf, flower, and fruit. The current study revealed the interplay between the fungal communities of the grape phyllosphere with the surrounding air. Plants were identified as a major source of recruitment for the atmospheric microbiome, while the surrounding atmosphere contributed only a small fraction of the plant fungal community. The results of the study suggested that the plant-air interface modulates the plant recruitment of atmospheric fungi, taking a step forward in understanding the plant holobiont assembly and how the atmosphere surrounding plants plays a role in this process. The impact of plants on the atmospheric microbiota has several biological and epidemiological implications for plants and humans.

Keywords: ITS amplicon libraries; aerobiology; fungal community; grapes; holobiont; metagenomics; microbiota; spore trap.

FIGURE 1

Network of shared and unique fungal OTUs among investigated samples. Colored edges (links) represent OTUs associated to sample types (air, fruit, leaves, and flowers). Light blue circle nodes connecting two or more samples represent shared OTUs. Numerical values indicate the number of OTUs and their percentage relative to the total number of OTUs.

FIGURE 2

Charts showing fractions of the estimated sources of fungal communities in each sink environment (air, leaves, and flowers/fruit) (A) and schematic description of the estimated source of the fungal communities associated with air and plant organs (B). The scheme shows the estimated movement of communities originated from leaves (green dashed arrows), flower (red dashed arrows), fruit (orange dashed arrows), and air (blue arrows). Dashed black circle arrows indicate the percentages of community commonly exchanged between two environments. Solid colored circle arrows showed the percentages of community that retune to the same environment from all the other sources (B).

FIGURE 3

Microbial association network showing interactions (co-occurrence and mutual exclusion) represented by green and red links, respectively. The size of the nods indicates the average relative abundance of OTUs across all investigated samples, while colored pie charts embedded inside the nodes show the relative distribution of each OTU in different samples (air, flowers, fruit, and leaves). Node shapes are used to differentiate fungal phyla. (A–D) are used to indicate the four sub-clusters that appeared to be influenced by the average relative abundance distribution of the interacting OTUs in the investigated samples. Interactions were calculated by CoNet, clustered using Connected Components algorithm as implemented in clusterMaker2, and visualized in Cytoscape 3.6.

FIGURE 4

Fungal taxa with significant temporal variations and a relative abundance of at least taxa 1% in different investigated samples (air, fruit, flowers, and leaves). The statistical comparison was done using a non-parametric Kruskal–Wallis test, and the p values were calculated through 1000 permutation and corrected using FDR method (Supplementary Table S2).