Thesis (MScAgric)--University of Stellenbosch, 2001. / ENGLISH ABSTRACT: Several attempts have been made to reduce Botrytis cinerea grey mould in vineyards and in storage by means of biological control. However, the so called "silver bullet" approach in utilising a single antagonist, has its limitations when compared with synthetic fungicides. Often the antagonist has a limited spectrum of activity and the duration of its effectiveness is less than that provided by synthetic fungicides. Furthermore, antagonists are more likely to be effective in preventing initial infection rather than resumption of latent infection. Therefore, due to the various infection sites in grape bunches utilised by B. cinerea and the fact that the pathogen can remain latent in the grapevine tissue, it may be possible to obtain effective control of the pathogen by integrating fungicides and different biological control agents each aimed at a different site in grape bunches, protecting the bunch at the various phenological stages of growth and under different micro climatic conditions. In this study the potential of three fungal antagonists (Glioc/adium roseum, Uloc/adium atrum and Trichoderma harzianum) and one yeast (Trichosporon pullulans) to colonise different sites in grape bunches, and to reduce B. cinerea infection, was investigated in commercial vineyards. As the biological control agents were used in an integrated system, the effect of various fungicides frequently applied to local vineyards on the organisms was also investigated. Fungicide trials were conducted taking into account two possible scenarios. Firstly, the possible effect of fungicides applied to the vineyard after an application of the biological control agent or shortly before the application of the biocontrol agent. This entailed exposing the biocontrol agents to relatively low concentrations of the active ingredient of the fungicides, similar to the residue levels to which these organisms would be exposed under field conditions. Secondly, the possibility of applying the organisms and the fungicides at the same time by making use of spray tank mixtures. This meant exposing the biocontrol agents to relatively high doses of the active ingredient of the various fungicides. Mycelial growth and germination tests were performed on agar in Petri dishes to determine the effect of fungicides. It was assumed that if the fungicide effectively inhibits the antagonist at 2.5 !-lg a.Uml, the fungicide and antagonist can not be used in an integrated programme. Based on this criterium, T harzianum can not be applied to vineyards with penconazole, mancozeb/metalaxyl, pyrifenox or mancozeb. In addition T harzianum can not be applied as tank mixtures with iprodione. However, T harzianum can be used in conjunction with pyrimethanil, folpan, iprodione, fosetyl-Al and copperhydroxide, provided the chemicals and the antagonist are applied alternately. Gliocladium roseum can not be applied in a tank mixture with pyrimethanil and penconazole, but can be used on grapevine in conjunction with penconazole, pyrifenox, pyrimethanil, iprodione and fosetyl-Al. Ulocladium atrum can not be applied with pyrimethanil and iprodione. Ulocladium atrum can be applied in conjunction with penconazole, pyrifenox, pyrimethanil, iprodione, fosetyl-Al and mancozeb. The fungus can be applied in a tank mixture with penconazole and pyrifenox. The antagonists were applied as conidial suspensions to bunches at various phenological stages in commercial vineyards planted with the wine grape cultivar Chardonnay in the Stellenbosch region, or the table grape cultivar Dauphine planted in Paarl region. Bunches were collected 2 wk after application, surface-sterilised and used for determining antagonist colonisation and B. cinerea infection at specific sites in the bunches. In Chardonnay, the antagonists colonised the different sites, but colonisation during the three seasons was inconsistent and sporadic. Ulocladium atrum and G. roseum colonised floral debris to a degree in the 1996 season. However, in the 1997 season these two antagonists did not develop from floral debris. Trichoderma harzianum colonised floral debris extensively in the 1996 season. In the 1997 season colonisation by T harzianum dropped, but unlike G. roseum and U atrum, T harzianum occurred at a low level in flowers. Ulocladium atrum only colonised bunches during bloom, and was not found in bunches monitored from pea-size stage to véraison. This finding suggests that the saprophyte colonised moribund and dead flower parts occurring in bunches during full bloom to the pre-pea size stage, and is not likely to be found in living tissue. Gliocladium roseum colonised grape berries and pedicels to some degree and T harzianum colonised these grape parts extensively. Botrytis cinerea occurred inconsistently and at low frequencies in the different sites in bunches. It was therefore not possible to comment on the effectivity of the various antagonists in the three seasons during which the trials were performed. However, it was noted that, during the peasize stage in 1996, when high levels of B. cinerea were recorded, T harzianum controlled these infections in the pedicels more effectively than any other treatment. / AFRIKAANSE OPSOMMING: ONDERDRUKKING VAN BOTRYTIS CINEREA DEUR ANTAGONISTE IN LEWENDE, AFSTERWENDE EN DOOIE WINGERDWEEFSEL Die benadering om Botrytis cinerea verrotting van wingerd met behulp van 'n enkele biologiese beheeragent in plaas van met sintetiese fungisiede te beheer, het sekere beperkinge. Antagoniste het dikwels 'n beperkte spektrum van aktiwiteit, en die duur van hul effektiwiteit is minder as dié van fungisiede. Antagoniste is gewoonlik ook minder effektief in die beheer van latente infeksie. Die patogeen het verder die opsie om druiwetrosse deur verskillende infeksieweë te koloniseer. Fungisiede kan druiwetrosse beter teen infeksie deur veelvuldige infeksieweë beskerm as 'n enkele antagonis. In die lig hiervan is die beheer van die patogeen deur 'n kombinasie van fungisiede en verskillende biologiese beheeragente, wat elk gemik is om 'n ander infeksiepunt in die druiwe te beskerm, ondersoek. Drie swamagtige antagoniste (Glioc/adium roseum, Uloc/adium atrum en Trichoderma harzianum) en een gis (Trichosporon pullulans) is in die ondersoek gebruik. Voorloper ondersoeke, waar twee moontlike scenarios in ag geneem is, is met fungisiede uitgevoer. In die eerste scenario is die effek van fungisiede, aangewend op wingerd kort vóór aanwending van die biologiese beheeragent, of kort ná aanwending, ondersoek. Hierdie proef het die blootstelling van die biologiese beheeragent aan relatief lae konsentrasies van die aktiewe bestanddeel van die fungisied, vergelykbaar met residuvlakke waaraan die organismes onder veldtoestande blootgestel sou word, behels. Tweedens is die moontlikheid om antagoniste en fungisiede gelyktydig as spuitpompmengsels toe te dien, ondersoek. In hierdie proef is die biologiese beheeragente aan relatief hoë dosisse van die aktiewe bestanddeel van verskillende fungisiede blootgestel. Miseliumgroei en ontkiemingstoetse is op agar in Petribakkies uitgevoer om die effek van die fungisiede te bepaal. As kriterium is aanvaar dat indien 'n fungisied die antagonis effektief by 2.5J..lglml aktiewe bestanddeel inhibeer, die fungisied en antagonis nie in 'n geïntegreerde program gebruik kan word nie. Gebaseer op hierdie kriterium kan T harnzianum nie aangewend word in 'n wingerd wat met penconazole, mancozeb/metalaxyl, pyrifenox of mancozeb behandel is nie. Ook kan T harzianum nie in 'n spuitpompmengsel met iprodione aangewend word nie. Trichoderma harzianum kan egter saam met pyrimethanil, folpan, iprodione en fosetyl-Al gebruik word, mits dié chemikalieë en die antagonis afwisselend aangewend word. Glioc/adium roseum kan nie in 'n spuitpompmengsel met pyrimethanil en penconazole aangewend word nie, maar kan saam met penconazole, pyrifenox, pyrimethanil, iprodione en fosetyl-Al gebruik word. Uloc/adium atrum kan nie saam met pyrimethanil, iprodione en fosetyl-Al gebruik word nie. Die swam kan wel in 'n spuitpompmengselmet penconazole en pyrifenox aangewend word. In verdere proewe is die antagoniste as spoorsuspensies op trosse op verskillende groeistadia in kommersiële wingerde, wat met die wyndruitkultivar Chardonnay of die tafeldruifkultivar Dauphine aangeplant is, ondersoek. Trossies is twee weke na toediening versamel, oppervlakkig gesteriliseer en gebruik om vlakke van antagoniskolonisasie en B. cinerea infeksie op spesifieke nisse in die trosse te bepaal. In die geval van Chardonnay het die antagoniste die verskillende nisse gekoloniseer, maar die kolonisasie was sporadies en nie konstant gedurende die drie seisoene van ondersoek nie. Uloc/adium atrum en G. roseum het blomdeeltjies tot 'n beperkte mate in die 1996 seisoen gekoloniseer, maar nie in die daaropvolgende seisoen nie. Daarteenoor het T. harzianum blomdeeltjies ekstensief in die 1996 seisoen gekoloniseer, en in 'n beperkte mate in die daaropvolgende seisoen. Uloc/adium atrum kon nie trosse van ertjiekorrelgrootte tot deurslaan vestig nie. Hierdie bevinding dui daarop dat die saprofiet afsterwende en dooie blomdeeltjies, wat van volblom tot ertjiekorrelstadium in die trosse voorkom, koloniseer, maar dat dit nie in lewende weefsel voorkom nie. Daarteenoor het T. harzianum die verskillende trosdele ekstensief gekoloniseer. Botrytis cinerea het gedurende die drie seisoene wisselvallig en teen lae frekwensies in die verskillende nisse in die trosse voorgekom. Dit was gevolglik nie moontlik om 'n konkrete afleiding oor die effektiwiteit van die verskillende antagoniste as biobeheeragente van B. cinerea te maak nie. In die geval van Dauphine was die onderskeie organismes swak koloniseerders van blomdeeltjies. Trichoderma harizanum kon egter die lewende trosdele koloniseer. Kolonisasievlakke was laag en was nooit meer as 50% nie. In beide seisoene het die kolonisasievermoë van T. harzianum drasties ná trostoemaak gedaal. Daarteenoor het beide G. roseum en U atrum tydens al die ontwikkelingstadia die lewende trosdele swak gekoloniseer. Botrytis cinerea het ook uiters sporadies en teen baie lae vlakke voorgekom. Die bevindinge het getoon dat klimaatsomstandighede wat in tafeldruifwingerde in die Wes-Kaap heers, nie geskik is vir die vestiging van die biologiese beheeragente wat in die studie ondersoek is nie.

This investigation has had for its primary purpose the noting of the incidence of antagonists as occurring in a Denton County, Texas, Norfolk Fine Sand. A further interest has been the determination of the effect of these antagonists upon known beneficial soil organisms.