Investigating the effect of leaf removal on the grape-associated microbiome through culture-dependent and –independent approaches

Morgan, Horatio Herbert (2016-12)

Thesis (MSc)--Stellenbosch University, 2016.

Thesis

ENGLISH ABSTRACT: Leaf removal is a common practice, which is often performed at either, veraison to increase the air circulation, light exposure, penetration of fungicides and decrease disease incidence; or at fruit-set to enhance wine colour, flavour and aroma. While there is greater understanding of how leaf removal influences the grape chemical composition and grapevine associated pathogenic fungi, most notably Botrytis cinerea, there is limited information on how the complete microbial community responds to these leaf removal procedures. The current study focused on a phylogenetic survey of both the bacterial and fungal communities present in Sauvignon blanc grape must prepared from shaded (SH) and sun-exposed (EX) grapes. A combination of culture-based methods, ARISA community fingerprinting as well as direct amplicon sequencing was employed to analyse the community. The yeast community was analysed using culture-based methods for four consecutive vintages (2012 – 2015), while the bacteria were only isolated and identified for the 2015 samples. Furthermore, the two culture-independent approaches were used to profile and identify the bacterial and fungal species present in the 2014 & 2015. A comparison of the cultivable yeast communities in the shaded and exposed grape derived must using ANOSIM revealed that there was more variation in the yeast community structures between vintages than between the treatments, thus suggesting the leaf removal only had a small effect on the yeast community (R = -0.04, p = 0.65). In contrast, ARISA data revealed a more diverse community which was different between the treatment (R = 0.5, p = 0.63) albeit with significant overlap. Our ARISA data suggested that leaf removal might have a significant influence on the filamentous fungal community and to a lesser extent on the yeast community. A more in-depth analysis of the microbiome was performed using Illumina target amplicon sequencing of the ITS-1 region, and the fungal population in the SH and EX musts exhibited significant overlap (R = -0.5, p =1) within treatments, while significant overlap between years was observed (R = 0.5, p =0.331) and the difference between the two populations was not significant. However, minor reductions in both Botrytis and Penicillium were observed in the sun-exposed derived grape must. The cultivable bacterial community derived from the 2015 must samples demonstrated complete variation between the samples, with the absence of AAB in the exposed derived must. However, this was only based on a single vintage. Contrastingly, the B-ARISA data demonstrated that leaf removal had no influence of on the bacterial community, with the variation greatest within replicate samples (treatment) (R = -1, p = 1). Our ARISA data suggested that vintage might have a greater influence in driving bacterial communities. The in-depth analysis, was achieved by targeting the protein encoding region rpoB as it has been demonstrated to provide greater resolution and overcome biases associated with the 16S rRNA region. The results were similar to that obtained in the fungal analysis (R = -0.5, p =1), with significant overlap within treatments with, some overlap between years (R = 0.25, p =0.64). However, potential plant pathogens Stenotrophomonas and Xanthomonas were reduced in the exposed treatment, together with Achromobacter and Bordetella. Interestingly concordant with the culture-dependent analysis, Gluconobacter was not identified in the 2015 exposed sample, but only in the shaded derived must. Overall our data show similar trends with regards to the microbial community composition in the shaded and exposed must and whether these are largely influenced by defoliation or not. As a whole the data set suggest that the differences in the microbial community can largely be ascribed to the absence and presence of minor species and relative abundance of a few major species, that dominate the berry surface. On the surfaces leaf removal appears to have no impact on the microbial community, however on closer inspection it seems as if a few groups are impacted, ever so slightly. Therefore, future work should focus on selectedmicrobial communities, which has been shown to be influenced by leaf removal or have the potential for disease. Moreover, the study did however only focus on a single variety, for two consecutive vintages, a greater sample set is required over several vintages and maybe two cultivars.

AFRIKAANS OPSOMMING: Blaar verwydering is ʼn algemene praktyk, wat gereeld uitgevoer word tydens of deurslaan om die lug vloei, lig blootstelling, deurdringing van plaagdoder te verhoog, en die voorkoms van siektes te verlaag, of tydens vrugset om wynkleur, geur en aroma te verhoog. Alhoewel daar ʼn groter begrip is van hoe blaar verwydering die druif chemiese samestelling en die wingerd geassosieerde patogeniese swamme, veral Botrytis cinerea, beïnvloed, is daar beperkte inligting oor hoe die algehele mikrobiese gemeenskap reageer op hierdie blaar verwyderings praktyke. Die huidige studie fokus op ʼn filogenetiese opname van beide die bakteriese en swam gemeenskappe teenwoordig in Sauvignon blanc druiwesap berei vanaf skadu (SH) en son-blootgestelde (EX) druiwe. ʼn Kombinasie van kultuur gebaseerde metodes, ARISA gemeenskap vingerafdruk sowel as direkte amplikon volgorde bepaling is toegepas om die gemeenskap te analiseer. Die gis gemeenskap is analiseer deur kultuur gebaseerde metodes op 4 agtereen volgende oesjare (2012-2015), terwyl die bakterieë slegs uit die 2015 oesjaar geïsoleer en identifiseer is. Verder is die twee kultuur-onafhanklike benaderings gebruik is om die bakteriese en swam spesies teenwoordig in 2014 en 2015 te identifiseer. ʼn Vergelyking van die kultiveerbare gis gemeenskappe in die skade en die blootgestelde druiwesap met behulp van ANOSIM het aan die lig gebring dat daar meer variasie in die gis gemeenskapstrukture tussen oesjare is as tussen die behandeling, wat daarop dui dat blaar verwydering slegs ʼn klein invloed het op die gis gemeenskap (R = -0.04, p = 0.65). In teenstelling onthul ARISA data ʼn meer diverse gemeenskap wat verskil tussen behandelinge (R = 0.5, p = 0.63) al is dit met beduidende oorvleueling. Ons ARISA data dui daarop dat blaar verwydering ʼn beduidende invloed het op die filamentagtige swam gemeenskap en tot ‘n mindere mate op die gis gemeenskap. ʼn Meer in diepte analise van die mikrobioom is gedoen met Illumina teiken amplikon volgorde bepaling van die ITS-1 streek, en dit wys dat die swam bevolkings in die SH en EX druiwemos beduidend oorvleuel (R = -0.5, p = 1) tussen behandelings, terwyl beduidende oorvleueling tussen oesjare waargeneem is (R = 0.5, p = 0.331) en die verskille tussen die twee bevolkings nie beduidend was nie. Alhoewel ʼn geringe afname in beide die Botrytis en Penicillium waargeneem is in die son-blootgestelde druiwemos. Die kultiveerbare bakteriese gemeenskap afkomstig van die 2015 mos monsters demonstreer algehele variasie tussen die monsters, met ʼn algehele afname in asynsuur bakterieë in die blootgestelde mos, maar die resultate is gebaseer op ʼn enkele oesjaar. Kontrasterend, wys die B-ARISA data dat blaar verwydering geen invloed gehad het op die bakteriese gemeenskap nie, met die grootste variasie tussen replikaat monsters (behandeling) (R = -1, p = 1). Ons ARISA data stel voor dat oesjaar ʼn groter invloed het op die bakteriese gemeenskappe. Die in-diepte analise is gedoen deur die proteïen koderende streek rpoB te teiken, aangesien dit ʼn groter resolusie vertoon en vooroordele wat verband hou met die 16S rRNA streek oorkom. Die resultate is soortgelyk aan die verkry vir die swam analise (R = -0.5, p = 1), met beduidende ooreenkomste tussen behandelings en ʼn mate van oorvleueling tussen oesjare (R = 0.25, p = 0.64). Potensiële plant patogene Stenotrophomonas en Xanthomonas verminder in die blootgestelde behandeling, tesame met Achromobacter en Bordetella. Interessantheidshalwe, met die kultuur-afhanklike ontleding, is Gluconobacter nie geïdentifiseer in die 2015 blootgestelde monster nie, maar slegs in die skadu afgeleide mos. Oor die algemeen wys ons data soortgelyke tendense met betrekking tot die mikrobiese gemeenskap samestelling in die skadu en son-blootgestelde mos, asook of hierdie gemeenskappe grootliks beïnvloed word deur blaar verwydering of nie. As ʼn geheel toon die data stel dat die verskille in die mikrobiese gemeenskap grootliks toegeskryf kan word aan die afwesigheid en teenwoordigheid van klein spesies en die relatiewe oorvloed van ʼn paar groot spesies, wat die bessie oppervlak oorheers. Op die oppervlak vertoon dit of blaar verwydering geen invloed het op die mikrobiese gemeenskap nie, maar by nadere ondersoek blyk dit dat ʼn paar groepe effens beïnvloed word. Daarom moet toekomstige werk fokus op geselekteerde mikrobiese gemeenskappe, wat getoon het dat dit deur blaar verwydering beïnvloed word of wat die potensiaal het om siektes te veroorsaak. Verder het die studie net gefokus op ʼn enkele kultivar vir twee opeenvolgende oesjare, ʼn groter monster stel is nodig oor verskeie oesjare en meer kultivars.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/100307
This item appears in the following collections: