Influence of vintage, geographic location and cultivar on the structure of microbial communities associated with the grapevine rhizosphere in vineyards of San Juan Province, Argentina

Abstract

Soil microbiomes, as a primary reservoir for plant colonizing fungi and bacteria, play a major role in determining plant productivity and preventing invasion by pathogenic microorganisms. The use of 16S rRNA and ITS high-throughput amplicon sequencing for analysis of complex microbial communities have increased dramatically in recent years, establishing links between wine specificity and, environmental and viticultural factors, which are framed into the elusive terroir concept. Given the diverse and complex role these factors play on microbial soil structuring of agricultural crops, the main aim of this study is to evaluate how external factors, such as vintage, vineyard location, cultivar and soil characteristics, may affect the diversity of the microbial communities present. Additionally, we aim to compare the influence these factors have on the structuring of bacterial and fungal populations associated with Malbec grapevine rhizosphere with that of the more widespread Cabernet Sauvignon grapevine cultivar. Samples were taken from Malbec and Cabernet Sauvignon cultivars from two different vineyards in the San Juan Province of Argentina. Total DNA extracts from the rhizosphere soil samples were sequenced using Illumina's Miseq technology, targeting the V3-V4 hypervariable 16S rRNA region in prokaryotes and the ITS1 region in yeasts. The major bacterial taxa identified were Proteobacteria, Bacteroidetes and Firmicutes, while the major fungal taxa were Ascomycetes, Basidiomycetes, Mortierellomycetes and a low percentage of Glomeromycetes. Significant differences in microbial community composition were found between vintages and vineyard locations, whose soils showed variances in pH, organic matter, and content of carbon, nitrogen, and absorbable phosphorus.

Fig 1. Vintage boxplot of microbial species Richness (S) and Shannon Index (H’) values.

Boxplots (A) and (B) represent the fungal and prokaryotic species richness variations, respectively. Correspondingly, figures (C) and (D) represent the fungal and prokaryotic Shannon Index estimate variations. Pairwise-comparisons between vintages for (S) and (H’) in both communities are also established. The statistical analysis was done separately for each community that included an initial (S) and (H’) comparison between all three vintages using a Kruskal Wallis test (Richness: p = 0.0024 for fungi and p = 0.0142 for prokaryotes; Shannon Index: p = 0.0323 for prokaryotes). The identified pairwise-comparisons results were obtained through a post hoc Mann–Whitney Bonferroni adjusted test.

Fig 2. Non-metric Multidimensional Scaling (NMDS) of microbial communities by vintage, location, cultivar and plots.

(A) Fungal communities grouped according to the vintage, (B) location, (C) cultivar and (D) plot. (E) Prokaryotic communities grouped according to the vintage, (F) location, (G) cultivar and (D) plot. Samples were grouped at 20% of dissimilarity distance using the generalized Unifrac metric. P values to observe significant difference between communities were obtained through a PERMANOVA test, and are represented on the left hand corner of each figure.

Fig 3. Average relative abundance of the most dominant microbial communities according to the vintage, vineyard location and cultivar.

Identified fungal communities classified by (A) vintage year, (B) vineyard location and by (C) cultivar. Identified prokaryotic communities’ classified by (D) vintage year (E) vineyard location and (F) cultivar. Only the communities with abundance > 0.1% are represented.

Fig 4. Canonical Correspondence Analysis (CCA) of fungal and bacterial communities.

(A) Represents the CCA analysis of the main abundant fungal classes and (B) represents the CCA analysis of the main abundant bacterial classes classified. The 2015 samples are marked in red, the 2016 in green and the 2017 in blue.