EDP Sciences logo
Web of Conferences logo
Search
Menu
All issues Volume 68 (2023) BIO Web Conf., 68 (2023) 02034 Full HTML
Open Access
Issue
BIO Web Conf.
Volume 68, 2023
44th World Congress of Vine and Wine
Article Number 02034
Number of page(s) 6
Section Oenology
DOI https://doi.org/10.1051/bioconf/20236802034
Published online 22 November 2023

© The Authors, published by EDP Sciences, 2023

Licence Creative CommonsThis is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1 Introduction

The microbial communities present in the grapes and the winery environment contribute to the diversity and complexity of the wine. Diverse and balanced microbiome on the grape berries can result in wines with more desirable flavours and aromas. As a result, winemakers carefully manage the microbiome during vinification. Traditional fermentation of wine typically involves the addition of a commercial yeast strain to the grape juice to initiate and control the fermentation process. Choice of yeasts for fermentation is dependent on the winemaker's preferences and the region of production. In contrast, spontaneous fermentation of wine relies on the naturally occurring microorganisms present in the grape must and winery environment to initiate and complete the fermentation process [1]. This approach can result in a more complex and unique flavour profile as a result of the diverse range of microorganisms involved [2]. However, spontaneous fermentation is more unpredictable and can be more challenging for winemakers to manage. One possible way to assess the composition and activity of microorganisms during fermentation is through the use of massive parallel sequencing technology [3]. Next generation sequencing platforms are producing millions of sequencing reads that need to be processed. Currently there are two main sequencing approaches to observe microbiome in the samples - amplicon based and whole genome shotgun sequencing, both with their own advantages and drawbacks [4]. Widely used tools for microbiome analysis based on amplicon sequencing are QIIME2 [5], mothur [6], and phyloseq [7]. Our aim is to test the pipeline for identification and quantification of microorganisms during fermentation, and how different starters affect other microorganisms during production of wine.

2 Material and methods

2.1 Sample collection

Ripe grapes of the Pinot blanc variety were harvested from the Little Carpathian wine-growing region, from the municipality of Modra, Slovakia in the years 2018, 2019, and 2020. Sulphur dioxide was added to fresh grape juice to concentration of 20 mg/ml free SO2. Tested batches were divided into 3 groups. First group was inoculated by commercially available S. cerevisiae (Ruland BS5, BS vinařské potřeby s. r. o., Czech Republic), second group was inoculated by L. thermotolerans and S. cerevisiae, and third group was inoculated with M. pulcherrima and S. cerevisiae. Each group consisted of 3 tanks (biological replicates). L. thermotolerans and M. pulcherrima were received from the National Agricultural and Food Centre, Food Research Institute. Inoculation was performed by adding 20 g of dried yeasts to a mixture of fresh grape juice and water in ratio 1:1 and heated up to 35 ˚C for 30 min and added to the rest of the must. The batches were mixed well and samples were taken from the bottom of the vessel for the different tested phases. Phases tested were: must appr. 2-3 days post inoculation (m1), actively fermenting must (m2), and young wine (wine before filtration; m3). Each sample was collected as 0.5 L of liquid except for phase m3 used for RNA isolation. For this purpose we collected 2 L of the liquid. Experimental schema is visualised in Fig. 1.

thumbnail Figure 1

Experimental and sampling scheme. Harvest from the same vineyard was collected for 3 years. It was divided into 3 groups: control - inoculated with commercial S. cerevisiae, lasa - inoculated with L. thermotolerans, and mesa - inoculated with M. pulcherrima. There were 3 biological replicates, totalling 9 tanks per year. Technical triplicates were taken from tanks. DNA and RNA samples were collected simultaneously. Created with BioRender.com

2.2 Nucleic acid isolation

DNA was isolated from 2 ml of must by DNeasy Mericon Food Kit (Qiagen, Hilden, Germany) according to the manufacturer instructions. For RNA isolation we used 50 ml of sample liquid that was centrifuged and the pellet was subsequently mixed with 10 ml of TRIzol (Thermo Fisher Scientific, Waltham, MA, USA) and frozen. RNA was extracted by the use of Nucleospin RNA Midi kit (Macherey-Nagel, Düren, Germany) and reversely transcribed to cDNA by Maxima H Minus First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, USA).

2.3 PCR amplification

Bacterial 16S rRNA was amplified using primer pair sequences for the V3 and V4 regions [8]. The primers were 16S amplicon PCR forward primer (5′-TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG CCT ACG GGN GGC WGC AG-3′) and 16S amplicon PCR reverse primer (5′-GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA GGA CTA CHV GGG TAT CTA ATC C-3′). Fungal 28S rRNA was amplified by 28S amplicon PCR forward primer (5′-TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG GAG TCG AGT TGT TTG GGA AT-3′) and 28S amplicon PCR reverse primer (5′-GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA GGG TCC GTG TTT CAA GAC GG-3′). All primers contained Illumina adapter regions.

The PCR mixture contained template DNA or cDNA (∼12 ng), 2x KAPA HiFi HotStart ReadyMix (12.5 μl; Kapa Biosystems, Wilmington, MA, USA) and amplicon primers (0.5 μl, 10 pM) in a total reaction volume of 25 μl. Reaction conditions consisted of an initial 95°C, 5 min; 35 cycles of 94°C for 60 s, 60°C for 16S (52°C for 28S), 45 s, and 72°C, 60 s, and final extension at 72°C, 7 min. After amplicon PCR, samples were purified using a QIAquick PCR Purification Kit (Qiagen, Hilden, Germany).

2.4 Sequencing

DNA libraries were prepared using the 16S Metagenomic Sequencing Library Preparation protocol (Illumina Inc, San Diego, CA, USA). Index PCR, a step that attaches dual indices and Illumina sequencing adapters using a Nextera XT Index kit (Illumina Inc, San Diego, CA, USA), was performed following the protocol using 2x KAPA HiFi HotStart ReadyMix (25 μl), Nextera XT Index Primers (5 μl per sample), PCR grade water and 5 μl of template DNA/cDNA solution in a total reaction volume of 50 μl. PCR Clean-Up 2 was performed as described in the protocol, followed by library validation using Qubit dsDNA high sensitivity assay (LifeTechnologies, Eugene, Oregon, USA) and Agilent® HS DNA Chip (Agilent® Technologies 2100 Bioanalyser). Finally, DNA libraries were normalised to 4 nM concentration, denatured and sequenced on a MiSeq system (Illumina Inc, San Diego, CA, USA) using a MiSeq Reagent kit v3 with paired-end of 2 × 300 bp reads.

2.5 Data analysis

After initial quality control of sequencing reads by FastQC [9], sequence data were processed in Qiime2 v2021.8 [5]. Raw reads were demultiplexed, filtered for quality, and sequencing primers were removed using cutadapt v3.4 [10]. From the processed reads, ASVs (amplicon sequence variants) were generated using DADA2 (via the q2-dada2 plugin) [11]. Individual ASVs were classified into taxonomy using the q2-feature-classifier plugin [12] which uses the classify-sklearn naïve Bayes algorithm, and the databases searched were Silva release 138.1 [13] for 16S rRNA, and RefSeq database (release 88) for 28S rRNA [14]. Sequences aligned to mitochondria and chloroplasts were filtered out. Computational analyses were written and executed using the SnakeLines framework [15, 16]. Visualisations were produced using R Statistical software (v4.3.0, [17]) and utilising package pals (v1.7, [18]).

3 Results

3.1 Analysis of fungi

Saccharomyces was the most prominent yeast contributing to the vinification (Fig. 2). Its role as major contributor to composition was superseded only by its own contribution to RNA activity where it was responsible for no less than 89.8% of total activity in the active fermenting (m2) phase. Hanseniaspora and Saccharomyces formed a major part of fungal composition with only exception for m1 phase in 2020, where inoculated Lachancea formed 27.7-86.5% of fungi in must. Hanseniaspora was present in the initial must and easily formed 90% of total composition in part of the samples during all examined phases (m1, m2, m3). Although Hanseniaspora was the second most abundant yeast during fermentation, its RNA activity was negligible with top at 4.6% of total RNA activity in initial must, but more commonly below 0.5% of total RNA activity. Hanseniaspora and Saccharomyces were the most abundant genera during all three years of experiments. Our experimental starter culture of Lachancea thermotolerans, which was a significant coloniser, shares similar inactivity at RNA level with Hanseniaspora. While it contributed to RNA activity 7.0% at the most, its usual activity was between 0.5-2.5% of total activity. Another significant eukaryotic microorganisms that were detected in some samples belonged to genera Pichia, Saturnispora, Aureobasidium, Erysiphe and Botrytis and were detected at levels ~3%, 1-5%, ~1%, ~1%, and ~0.5%, respectively. Saturnispora was primarily detected in the phase m3 in the years 2019 and 2020. Aureobasidium and Erysiphe were detected primarily in phase m1 in 2018 and 2019. Botrytis was detected mostly in the m1 phase in 2018. Aureobasidium was detected almost exclusively with Erysiphe, and Botrytis was paired with Alternaria. What is interesting is that these plant pathogens show significant activity at RNA level during the m1 phase in 2018.

We tested two potential non-S. cerevisiae starter cultures for inoculation - Lachancea thermotolerans and Metschnikowia pulcherrima. With Lachancea we were successful in colonising the must, but for Metschnikowia we did not observe successful colonisation. Metschnikowia was detected at levels well below 0.5% of total DNA. Highest amount detected was 2.4-3.4% in initial must at 2018, but that must was not inoculated and Metschnikowia was probably of an indigenous nature.

thumbnail Figure 2

Fungal composition and activity profile. Only genera that accounted for at least 1% on average across all samples or exceeded 5% in individual samples were included in the visualisation. Eleven genera collectively formed more than 97% of the entire composition and activity. Data shown for single biological replicate from each year.

3.2 Analysis of bacteria

Bacterial compositional changes were less dynamic compared to the fungi (Fig. 3). Richness of bacterial populations was highest in phase m1 followed by m3 and lowest during phase m2. As fermentation proceeded, both mitochondria and chloroplasts exhibited a gradual reduction at the DNA and RNA levels. Amplicons of 16S rRNA genes that were classified as mitochondrial or chloroplastic (remnants of grapevine cells) were discarded from this analysis. Looking at phylum level, Pseudomonadota absolutely dominated the composition of wine bacteria. Other phyla present in decreasing order of abundance were Bacillota, Actinomycetota and Bacteroidota. In rare cases, the composition of Bacillota reached up to 59% of the total sequences. However, more commonly, it occupied 5-22% of the bacterial population, with a preference for the later phases of fermentation. Actinomycetota and Bacteroidota were present in roughly the same amounts and formed 3-6% of composition. Moving to genera level, the most abundant genus was Gluconobacter followed by Komagataeibacter, Pseudomonas and Acetobacter. Gluconobacter, Komagataeibacter and Acetobacter collectively formed 40-80% of total bacterial composition. Gluconobacter was the most active genus at RNA level producing 15-52% of total activity. Noticeable is 65-87% share of RNA activity by E. coli in all phases of fermentation in 2018. Genera Variovorax, Massilia, Tatumella and Sphingomonas were also extensively detected in samples. In solitary samples we detected high activity of Oceanobacillus, Methylorubrum, Commensalibacter and Aquabacterium. Oenococcus was present only in minute quantities and in a few samples. It was probably replaced by other lactic acid producing bacteria (LAB), especially Lactobacillus, Lactococcus and Streptococcus. Generally LAB were in low quantities especially when compared to bacteria producing acetic acid (AAB). Only exception is in 2020, where Lactobacillus formed 4-24% of the bacterial population during phases m1 and m2. Even higher was Lactobacillus activity at RNA level with 21-74% of total activity observed in two different tanks. Presence of Lactobacillus was in negative correlation with the high abundance of Pseudomonas. There seems to be no link between specific bacteria and tested starters.

thumbnail Figure 3

Bacterial composition and activity profile. Only genera that accounted for at least 1% on average across all samples or exceeded 5% in individual samples were included in the visualisation. Twenty genera collectively formed more than 69% of the entire composition and activity. Abbreviations o. and f. are for taxa that were classified only to the rank of order and family. Data shown for single biological replicate from each year.

4 Discussion

When assessing the effectiveness of the metagenomics pipeline, it is crucial to acknowledge and discuss the limitations of our study. First and the most important is that we are analysing sequence abundance, not true taxonomic abundance [19]. This difference is rooted in a different copy number of rRNA genes per taxon. Copy number variation can be quite variable even within a species, hence sequence estimation is best we can do by this approach. Second limitation is that we use only V3-V4 regions of the 16S rRNA gene and ITS1 region for taxonomic classification. To mitigate this inaccuracy we chose to infer taxonomy on genus level. That being said, we were able to observe complex composition and activity during the whole fermentation process.

It was expected and confirmed that Saccharomyces is the most important microbe in wine production and dominates composition and activity above all other microbes. In our experimental set up, addition of SO2 to the must favours the resistant S. cerevisiae or Saccharomycodes ludwigii at the expense of more sensitive non-Saccharomyces species, such as Kloeckera [20]. Morgan et al. reported that addition of sulphur dioxide at crush did not impact bacterial composition [21]. In general, the inoculated strain is primarily responsible for fermentation, but that the indigenous strains are not suppressed during the first several days of fermentation, and hence they may play a significant role [20]. In our case traditional inoculation with S. cerevisiae was highly effective when compared to two tested starters. Lachancea thermotolerans formed part of yeast population during beginning of fermentation, but its activity at RNA level was minimal and no significant impact is expected on flavour profile. Metschnikowia pulcherrima wasn’t inoculated successfully. Biggest differences in yeast profiles were observed between seasons.

In 2018 we could see there are several plant pathogens Erysiphe, Botrytis and Penicillium in all phases of fermentation. Erysiphe is a common plant pathogen causing powdery mildew [22]. Botrytis might be causing noble rot on overripen grapes in specific environmental conditions and is used to produce botrytic wines like those from Tokaj region [20]. This is probably not the case as it is present only in a small number of samples going through all phases. We assume that it comes from bunch rot on part of the harvest in that year. Penicillium is also considered a food spoilage microorganism causing moulds [23]. Spoilage was probably so severe that Aureobasidium started to grow in pair with Erysiphe as it is natural biocontrol agent for Aspergillus and other grape vine pathogenic fungi [24]. Striking was an extensive activity of Escherichia in all phases of fermentation in 2018, and is a clear sign of contamination of grapes. The elevated presence of microbial spoilage organisms in wine likely shares a common cause, which can be attributed to significant precipitation and prolonged periods of heat during the harvest season. During the period prior to harvest, in September 2018, the cumulative precipitation was 30 mm higher than the average of the past 30 years. Additionally, the daily mean temperature was 4 °C above the average for the last 20 days leading up to the harvest [25]. Warm and damp weather during harvest should be indicative for additional precautions against bacterial and fungal pests.

Bacterial composition seems to be dependent on the type of wine produced [26, 27]. The abundance of AAB in our bacterial profile was comparable to that observed in a Riesling wine [26]. The observed AAB (specifically Gluconobacter, Komagataeibacter, and Acetobacter) are known to produce acetic acid through the degradation of ethanol. Additionally, Gluconobacter has the ability to convert leftover glucose, produced by yeasts, into gluconic acid. The AAB are not generally a problem as long as the fermenting must and the wines are kept totally anaerobic. During active fermentation, carbon dioxide is being rapidly produced, and it blankets the must, keeping it anaerobic [28]. While high levels of AAB are typically considered a spoilage issue in wine, in the production of white wine, a certain level of AAB and the resulting acidity are actually desired, especially when compared to the preferences for acidity in red wine varieties. Pseudomonas is considered spoilage organism in food, but it was detected in several studies in high abundances as part of bacterial microbiome in wine [27,29,30]. Oenococcus was naturally replaced by other LAB (specifically Lactobacillus, Lactococcus and Streptococcus) in our experiment. Although malolactic fermentation in white wines is not common this substitution might lead to different organoleptic characteristics [31]. Tatumella was frequently detected in high abundances, often being the most abundant bacterium, across different wine types [26]. However, in our study, we observed its consistent presence but at a relatively low level. This finding aligns with previous observations that higher abundance of Tatumella is negatively correlated with total acid content [26]. It is noteworthy that the majority of the bacteria detected in our study belonged to AAB.

In future studies, it is important to focus on the investigation of the relationship between the wine microbiome and the volatile compounds that contribute to desirable characteristics.

5 Conclusion

Massive parallel DNA sequencing technology and pipelines for processing such data are able to describe complex communities during the whole process of wine production. It enables wine producers to experiment with new and exotic cultures and observe how these experimental cultures affect microbes during all stages of fermentation. As we presented, metagenomic data acquisition and analysis is ideal both for controlled production and for spontaneous fermentation that’s becoming a trend in some wineries. The main advantage is in selection of biomarkers for mitigation of problematic batches or pinpoint which strains to supplement. Applied metagenomics can improve the description of the biotic factor of terroir and offers data-driven decisions in wine production.

This publication is the result of support from the Operational Programme Integrated Infrastructure for the projects: ITMS: 313011ATL7 (PanClinCov), ITMS: 313011V578 (PreveLynch), and ITMS: 313021BUZ3 (USCCCORD), co-financed by the European Regional Development Fund.

References

  • H. Piao et al., ‘Insights into the bacterial community and its temporal succession during the fermentation of wine grapes’, Front. Microbiol. 6, Aug. 2015, doi:10.3389/fmicb.2015.00809 [CrossRef] [Google Scholar]
  • D. Liu, J.-L. Legras, P. Zhang, D. Chen, and K. Howell, ‘Diversity and dynamics of fungi during spontaneous fermentations and association with unique aroma profiles in wine’, Int. J. Food Microbiol. 338, 108983, Jan. 2021, doi:10.1016/j.ijfoodmicro.2020.108983 [CrossRef] [Google Scholar]
  • M. Böhmer et al., ‘Comparison of microbial diversity during two different wine fermentation processes’, FEMS Microbiol. Lett. 367, 18, 150, Sep. 2020, doi:10.1093/femsle/fnaa150 [CrossRef] [Google Scholar]
  • I. Stefanini and D. Cavalieri, ‘Metagenomic Approaches to Investigate the Contribution of the Vineyard Environment to the Quality of Wine Fermentation: Potentials and Difficulties’, Front. Microbiol. 9, 991, May 2018, doi:10.3389/fmicb.2018.00991 [CrossRef] [Google Scholar]
  • E. Bolyen et al., ‘Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2’, Nat. Biotechnol. 37, 8, 852857, Aug. 2019, doi:10.1038/s41587-019-0209-9 [Google Scholar]
  • P.D. Schloss et al., ‘Introducing mothur: Open-Source, Platform-Independent, Community-Supported Software for Describing and Comparing Microbial Communities’, Appl. Environ. Microbiol. 75, 23, 7537–7541, Dec. 2009, doi:10.1128/AEM.01541-09 [CrossRef] [PubMed] [Google Scholar]
  • P.J. McMurdie and S. Holmes, ‘phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data’, PLoS ONE 8, 4, e61217, Apr. 2013, doi:10.1371/journal.pone.0061217 [CrossRef] [PubMed] [Google Scholar]
  • A. Klindworth et al., ‘Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies’, Nucleic Acids Res. 41, 1, e1–e1, Jan. 2013, doi:10.1093/nar/gks808 [CrossRef] [PubMed] [Google Scholar]
  • S. Andrews, ‘FastQC: A Quality Control Tool for High Throughput Sequence Data’. 2010. [Online]. Available: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ [Google Scholar]
  • M. Martin, ‘Cutadapt removes adapter sequences from high-throughput sequencing reads’, EMBnet.journal 17, 1, 10, May 2011, doi:10.14806/ej.17.1.200 [CrossRef] [Google Scholar]
  • B.J. Callahan, P.J. McMurdie, M.J. Rosen, A.W. Han, A.J.A. Johnson, and S.P. Holmes, ‘DADA2: High-resolution sample inference from Illumina amplicon data’, Nat. Methods 13, 7, 581–583, Jul. 2016, doi:10.1038/nmeth.3869 [CrossRef] [PubMed] [Google Scholar]
  • N.A. Bokulich et al., ‘Optimizing taxonomic classification of marker-gene amplicon sequences with QIIME 2’s q2-feature-classifier plugin’, Microbiome 6, 1, 90, Dec. 2018, doi:10.1186/s40168-018-0470-z [CrossRef] [PubMed] [Google Scholar]
  • C. Quast et al., ‘The SILVA ribosomal RNA gene database project: improved data processing and web-based tools’, Nucleic Acids Res. 41, D1, D590–D596, Nov. 2012, doi:10.1093/nar/gks1219 [CrossRef] [PubMed] [Google Scholar]
  • N.A. O’Leary et al., ‘Reference sequence (RefSeq) database at NCBI: current status, taxonomic expansion, and functional annotation’, Nucleic Acids Res. 44, D1, D733–D745, Jan. 2016, doi:10.1093/nar/gkv1189 [CrossRef] [PubMed] [Google Scholar]
  • J. Budis et al., ‘SnakeLines: integrated set of computational pipelines for sequencing reads’, 2021, doi:10.48550/ARXIV.2106.13649 [Google Scholar]
  • J. Köster and S. Rahmann, ‘Snakemake—a scalable bioinformatics workflow engine’, Bioinformatics 28, 19, 2520–2522, Oct. 2012, doi:10.1093/bioinformatics/bts480 [CrossRef] [PubMed] [Google Scholar]
  • R Core Team, ‘R Statistical software’. 2023. [Online]. Available: https://www.R-project.org/ [Google Scholar]
  • K. Wright, ‘Pals: Color Palettes, Colormaps, and Tools to Evaluate Them’. 2021. [Online]. Available: https://kwstat.github.io/pals/ [Google Scholar]
  • Z. Sun et al., ‘Challenges in benchmarking metagenomic profilers’, Nat. Methods 18, 6, 618–626, Jun. 2021, doi:10.1038/s41592-021-01141-3 [CrossRef] [PubMed] [Google Scholar]
  • P.S. Nigam, ‘WINES | Production of Special Wines’, in Encyclopedia of Food Microbiology, Elsevier, 2014, 793–799. doi:10.1016/B978-0-12-384730-0.00358-X [CrossRef] [Google Scholar]
  • S.C. Morgan, M. Tantikachornkiat, C.M. Scholl, N.L. Benson, M.A. Cliff, and D.M. Durall, ‘The effect of sulfur dioxide addition at crush on the fungal and bacterial communities and the sensory attributes of Pinot gris wines’, Int. J. Food Microbiol. 290, 1–14, Feb. 2019, doi:10.1016/j.ijfoodmicro.2018.09.020 [CrossRef] [Google Scholar]
  • H.-H. Kassemeyer and B. Berkelmann-Löhnertz, ‘Fungi of Grapes’, in Biology of Microorganisms on Grapes, in Must and in Wine, H. König, G. Unden, and J. Fröhlich, Eds., Berlin, Heidelberg: Springer Berlin Heidelberg, 2009, 61–87. doi:10.1007/978-3-540-85463-0_4 [CrossRef] [Google Scholar]
  • G. G. Khachatourians, ‘Biochemical and Modern Identification Techniques | Food Spoilage Flora’, in Encyclopedia of Food Microbiology, Elsevier, 2014, 244–249. doi:10.1016/B978-0-12-384730-0.00035-5 [CrossRef] [Google Scholar]
  • M. Dimakopoulou et al., ‘Phyllosphere grapevine yeast Aureobasidium pullulans reduces Aspergillus carbonarius (sour rot) incidence in wine-producing vineyards in Greece’, Biol. Control 46, 2, 158–165, Aug. 2008, doi:10.1016/j.biocontrol.2008.04.015 [CrossRef] [Google Scholar]
  • SHMU, ‘Slovak Hydrometeorological Institute: Historical precipitation data’. 2023. [Online]. Available: https://www.shmu.sk/en/?page=1&id=klimat_operativneudaje2&identif=11816&rok=2018&obdobie=1981-2010 [Google Scholar]
  • A.M. Bubeck et al., ‘Bacterial microbiota diversity and composition in red and white wines correlate with plant-derived DNA contributions and botrytis infection’, Sci. Rep. 10, 1, 13828, Aug. 2020, doi:10.1038/s41598-020-70535-8 [CrossRef] [Google Scholar]
  • S. Zhang, X. Chen, Q. Zhong, X. Zhuang, and Z. Bai, ‘Microbial Community Analyses Associated with Nine Varieties of Wine Grape Carposphere Based on High-Throughput Sequencing’, Microorganisms 7, 12, 668, Dec. 2019, doi:10.3390/microorganisms7120668 [CrossRef] [PubMed] [Google Scholar]
  • K.H. Steinkraus, ‘Wine’, in Encyclopedia of Microbiology, Elsevier, 2009, 138–143. doi:10.1016/B978-012373944-5.00127-9 [CrossRef] [Google Scholar]
  • C. Pinto et al., ‘Wine fermentation microbiome: a landscape from different Portuguese wine appellations’, Front. Microbiol. 6, Sep. 2015, doi:10.3389/fmicb.2015.00905 [CrossRef] [Google Scholar]
  • N.A. Bokulich, J.H. Thorngate, P.M. Richardson, and D.A. Mills, ‘Microbial biogeography of wine grapes is conditioned by cultivar, vintage, and climate’, Proc. Natl. Acad. Sci. 111, no. 1, Jan. 2014, doi:10.1073/pnas.1317377110 [CrossRef] [Google Scholar]
  • C. Virdis, K. Sumby, E. Bartowsky, and V. Jiranek, ‘Lactic Acid Bacteria in Wine: Technological Advances and Evaluation of Their Functional Role’, Front. Microbiol. 11, 612118, Jan. 2021, doi:10.3389/fmicb.2020.612118 [CrossRef] [Google Scholar]

All Figures

thumbnail Figure 1

Experimental and sampling scheme. Harvest from the same vineyard was collected for 3 years. It was divided into 3 groups: control - inoculated with commercial S. cerevisiae, lasa - inoculated with L. thermotolerans, and mesa - inoculated with M. pulcherrima. There were 3 biological replicates, totalling 9 tanks per year. Technical triplicates were taken from tanks. DNA and RNA samples were collected simultaneously. Created with BioRender.com
In the text
thumbnail Figure 2

Fungal composition and activity profile. Only genera that accounted for at least 1% on average across all samples or exceeded 5% in individual samples were included in the visualisation. Eleven genera collectively formed more than 97% of the entire composition and activity. Data shown for single biological replicate from each year.
In the text
thumbnail Figure 3

Bacterial composition and activity profile. Only genera that accounted for at least 1% on average across all samples or exceeded 5% in individual samples were included in the visualisation. Twenty genera collectively formed more than 69% of the entire composition and activity. Abbreviations o. and f. are for taxa that were classified only to the rank of order and family. Data shown for single biological replicate from each year.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.