Vineyard Soil Microbiome Composition Related to Rotundone Concentration in Australian Cool Climate 'Peppery' Shiraz Grapes

Abstract

Soil microbial communities have an integral association with plants and play an important role in shaping plant nutrition, health, crop productivity and product quality. The influence of bacteria and fungi on wine fermentation is well known. However, little is known about the role of soil microbes, other than microbial pathogens, on grape composition or their role in vintage or site (terroir) impacts on grape composition. In this study, we used an amplicon sequencing approach to investigate the potential relationships between soil microbes and inherent spatial variation in grape metabolite composition - specifically, the concentration of the 'impact aroma compound' rotundone in Shiraz grapes (Vitis vinifera L.) grown in a 6.1 ha vineyard in the Grampians region of Victoria, Australia. Previous work had demonstrated temporal stability in patterns of within-vineyard spatial variation in rotundone concentration, enabling identification of defined 'zones' of inherently 'low' or 'high' concentration of this grape metabolite. 16S rRNA and ITS region-amplicon sequencing analysis of microbial communities in the surface soils collected from these zones indicated marked differences between zones in the genetic diversity and composition of the soil bacterial and fungal microbiome. Soils in the High rotundone zone exhibited higher diversity of bacteria, but lower diversity of fungi, compared to the soils in the Low rotundone zone. In addition, the network analysis of the microbial community in the High rotundone zone soils appeared well structured, especially with respect to the bacterial community, compared to that in the Low rotundone zone soils. The key differences in the microbial community structure between the rotundone zones are obvious for taxa/groups of both bacteria and fungi, particularly for bacteria belonging to Acidobacteria-GP4 and GP7, Rhizobiales, Gaiellaceae, Alphaproteobacteria and the Nectriaceae and Tremellaceae families of fungi. Although mulching in some parts of the vineyard caused changes in bacterial and fungal composition and overall microbial catabolic diversity and activity, its effects did not mask the rotundone zone-based variation. This finding of a systematic rotundone zone-based variation in soil microbiomes suggests an opportunity to bring together understanding of microbial ecology, plant biochemistry, and viticultural management for improved management of grape metabolism, composition and wine flavor.

Keywords: bacteria; fungi; grapes; microbiome diversity; rotundone.

FIGURE 1

Soil sampling locations, rotundone zones and the mulched strip in the 6.1 ha study vineyard. The rotundone zones of Bramley et al. (2017) were identified through k-means clustering of map layers of grape berry rotundone concentration in 2012 (’12), 2013 (’13), and 2015 (’15) along with apparent electrical soil conductivity (EC_a), slope (Sl) and aspect, expressed in terms of orientation away from north (fN). The values in the legend are the zone means. Note that the overall block mean berry rotundone concentration varied by a factor of 40-fold over the 3 years in which rotundone concentrations were mapped. Bramley et al. (2017) provide further details.

FIGURE 2

Diversity and abundance of bacteria (16S rRNA) (A) and fungi (ITS region) (B) in surface (0–15 cm) soils from the three rotundone-based zones. Bars with different alphabets are significantly different from each other at P < 0.05.

FIGURE 3

Venn diagram showing the number of unique OTUs of (A) bacteria and (B) fungi for each of the rotundone zone and shared between the three rotundone-based soil zones.

FIGURE 4

Composition of bacterial communities in the surface soils in the different rotundone-based soil zones. (A) Canonical analysis of principle (CAP) ordination, constrained by zone; relative abundances (B) at phylum level, (C–E) at family level within selected phyla showing significant variation between zones.

FIGURE 5

Composition of fungal communities in the surface soils in the different rotundone-based soil zones. Panel (A) at Phylum level, (B) distance based redundancy analysis (dbRDA); relative abundances (C) at Class level and (D) at family level showing variation between zones and depths.

FIGURE 6

Key indicator bacterial (A) and fungal (B) groups significantly different between the Low and High rotundone zones.

FIGURE 7

Composition of bacterial and fungal communities in the surface soils in the different rotundone-based soil zones. Canonical analysis of principle (CAP) ordination for (A) Bacteria and (B) Fungi, constrained by zone based on the Bray–Curtis similarity distance metrics. Vectors in the circle represent fitted values of soil properties showing Pearson correlations r > 0.50.

FIGURE 8

Comparison of rotundone-Low (A) and rotundone-High without mulch (B) bacterial networks. Circles represent nodes whose size indicates connectivity, node color represents taxonomy at the phyla level. Edges indicate co-occurrence between nodes colored either blue for positive or red for negative. Each circular grouping is a module. Numbers within modules correspond to numbers indicated in the hierarchical clustering. (C) Hierarchical clustering based on Pearson correlations among module-eigengenes and a heatmap of module eigengenes of the rotundone-High without mulch network.

FIGURE 9

Comparison of rotundone-Low (A) and rotundone-High without mulch (B) ITS networks. Circles represent nodes whose size indicates connectivity, node color represents taxonomy at the phyla level. Edges indicate co-occurrence between nodes colored either blue for positive or red for negative. Each circular grouping is a module. Numbers within modules correspond to numbers indicated in the hierarchical clustering. Hierarchical clustering based on Pearson correlations among module-eigengenes and a heatmap of module eigengenes of the (C) rotundone-Low network (D) rotundone-High without mulch network.

FIGURE 10

Microbial catabolic profiling analysis results for the surface soil samples from the rotundone-zone soil samples. (A) Canonical variate analysis (CVA) plot showing the dissimilarity in the catabolic diversity of soil microbial communities, (B) heat map of the AWCD showing differences in substrate use efficiency for the various C-substrates by the soil microbial communities.