Thesis (PhD(Agric))--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: The purpose of this study was to characterize the race and vegetative compatibility of Fusarium oxysporum f. sp. melonis (FOM) isolates collected in the major melon producing areas, to report on their geographical distribution, and their possible relatedness to isolates from other countries. Seventy two FOM isolates obtained from 30 fields in 17 melon producing regions were race-typed using the differential cultivars Topmark (susceptible to all races), Doublon (Fomi), CM 17187 (Fom2) and Perlita (Fom3) and grouped by means of vegetative compatibility. All isolates belonged to vegetative compatibility group 0134, indicating a high degree of genetic homogeneity among the South African FOM population. Fifty four isolates were identified as race 0, eight as race 1, and 10 as race 2. Race 0 occurred in 15 of the regions whereas race 1 was sporadically recovered. Race 2, on the other hand, was obtained only from four fields located in one geographical region. Perlita plants (carrying the gene Fom3) inoculated with local isolates ofrace 0 and race 2 and reference isolates of race 0 became stunted, their leaves turned yellow, and became thickened and brittle. These results suggested that Fom3 in Perlita confers a tolerant reaction compared to the resistant reaction of gene FornI in Doublon. The disease reaction of cultivar Perlita to FOM was therefore reinvestigated. Twenty isolates, including the four FOM races (0, 1, 2, and 1,2) obtained from different countries, were used. The differential cultivars were included to verify virulence of the isolates. Perlita plants inoculated with three isolates of race 2 remained asymptomatic. The remaining race 2 and 0 isolates, induced severe stunting of Perlita plants, but mean percentage stunting values did not differ significantly (P = 0.05) and ranged between 25.1 and 50.0. Leaves of stunted plants were chlorotic, thickened and brittle. Disease reaction of Perlita was verified at a lower inoculum concentration with two race 2 (pipette method) and two race 0 isolates (root dip method). Results proved that Fom3 does not confer similar resistance towards race 0 and some race 2 isolates as FornI in Doublon. Cultivars possessing Fom3, should therefore be considered tolerant to FOM races 0 and 2. The ability of a nit mutant isolate, generated from FOM race 0 which belongs to VCG 0134, to change its virulence during infection of melon plants, was investigated under quarantine. Seedlings of melon cultivars Imperial 45 and Early Sweet (no resistance genes), Amber (Fom2) and Fiata (FomI, Fom2) were consecutively grown in two cement troughs in a gauzehouse. Each planting was terminated when plants had advanced Fusarium wilt or after the fruit were harvested. In the first planting, Imperial 45 seedlings were transplanted and artificially inoculated with the nil mutant isolate. In the consecutive plantings, seeds were sown in the infested soil to enable natural infection. For each crop, representative plants showing Fusarium wilt were selected for isolation. All F. oxysporum isolates recovered were single-spored and their nit mutant and VCG status verified. Virulence of the labelled isolates was determined using differential cultivars. In trough A, all plants of the susceptible cultivars Imperial 45 and Early Sweet crops showed Fusarium wilt. The labelled isolates recovered from the selected plants were all designated race O. In the first crop (planting No.5) of the resistant cultivar Amber, 6.7% of the plants developed Fusarium wilt. In the second Amber crop the disease incidence increased to 56.6%, and to 81.8% in the final crop. Contrary to the susceptible cultivars, only race 2 isolates were obtained from the symptomatic Amber plants. Similar data were found with the susceptible cultivar Imperial 45 and the resistant cultivar Amber in trough B. Planting of Fiata caused a dramatic reduction in Fusarium wilt incidence in trough B. However, 1.2% of plants were affected by Fusarium wilt in the first Fiata crop (planting No.6), whereas 4% of the plants were symptomatic in the final planting. From these symptomatic Fiata plants only race 1,2 isolates were obtained. These findings, and the fact that the symptomatic plants represented a substantial proportion of the first Amber (approximately 7-15%) and Fiata (approximately 2%) crops, provedthat changes in the race structure of this fungal pathogen occurred rapidly when confronted with a resistant cultivar. The potential of RAPD analysis to differentiate between the isolates displaying virulence changes was evaluated. Four F. oxysporum f. sp. niveum isolates were included as an outgroup. A histopathological study was conducted to verify whether these isolates retain their ability to behave as true vascular pathogens. The three primers used clearly distinguished the 12 FOM isolates from the four F. oxysporum f. sp. niveum isolates. However, the primers showed a highly conserved and characteristic banding pattern for the FOM isolates which represented three physiological races (race 0, race 2, race 1,2), indicating that RAPD analysis cannot detect race-specific groupings in FOM. Disease reactions on the three differential cultivars confirmed the virulence of FOM isolates. The histopathological data furthermore proved that the two FOM races (race 2, race 1,2), which derived from the race 0 parent isolate, retained their ability to behave as true vascular pathogens. / AFRIKAANSE OPSOMMING: DIE KARAKTERISERING EN PATOGENESITEIT VAN SUID-AFRIKAANSE ISOLATE VAN FUSARIUMOXYSPORUMF. SP.MELONIS Die doel van die studie was om Fusarium oxysporum f. sp. melonis (FOM) isolate wat in die hoof spanspekproduserende gebiede versamel is, volgens ras en vegetatiewe verenigbaarheid te karakteriseer, en hul geografiese verspreiding en verwantskap met isolate van ander lande aan te dui. Twee en sewentig FOM isolate afkomstig vanaf 30 landerye wat 17 spanspekproduserende areas verteenwoordig, is gebruik. Die differensiële kultivars Topmark (vatbaar vir alle rasse), Doublon (Forni), CM 17187 (Fom2) en Perlita (Fond) is gebruik om die rasbepalings te doen asook om die vegetatiewe verenigbare groepe (VVG) te bepaal. Al die isolate is as VVG 0134 geklassifiseer, wat 'n hoë mate van genetiese homogenesiteit binne die Suid-Afrikaanse populasie aandui. Vier en vyftig isolate is as ras 0, agt as ras 1 en 10 as ras 2 geklassifiseer. Ras 0 is vanaf 15 gebiede afkomstig, terwyl ras 1 sporadies voorgekom het. Ras 2 is vanuit vier landerye binne dieselfde geografiese gebied verkry. Plante van die kultivar Perlita wat met plaaslike isolate van ras 0 en 2, asook verwysings-isolate van ras 0 geïnokuleer is, het verdwerg voorgekom. Die blare van die plante het vergeel, verdik en bros voorgekom. Hierdie siekte reaksie het aangedui dat Fond in Perlita toleransie bewerkstellig in teenstelling met die weerstandbiedende reaksie van geen Fomi in Doublon. Die siekte reaksie van Perlita teenoor FOM is dus verder ondesoek. Hiervoor is 20 isolate wat al vier FOM rasse insluit (0, 1, 2, en 1,2), en van verskillende wêrelddele afkomstig is, gebruik. Die virulensie van die isolate is met die differensiële kultivars bevestig. Drie van die ras 2 isolate het geen siektesimptome op Perlita veroorsaak nie. Die ander ras 2 isolate, en al die ras 0 isolate, het egter die Perlita plante aansienlik verdwerg en die blare vergeel en verdik. Laasgenoemde groep isolate het 'n gemiddelde verdwergingsindeks van tussen 25.1% en 50.0% veroorsaak. Die siekte reaksie by Perlita is verder bevestig deur plante teen 'n laer inokulumdigtheid van twee ras 2 (pipet metode), en twee ras 0 (wortel-doop metode) isolate, te inokuleer. Uit die resultate was dit duidelik dat die weerstand wat Fom3 teenoor ras 0 en sommige ras 2 isolate verskaf, van FornI verskil. Kultivars wat oor die weerstandsgeen Fom3 beskik moet dus as tolerant beskou word. 'n Ondersoek is geloods na die vermoë van 'n nil mutant isolaat, genereer vanaf die wilde ras 0 isolaat van FOM (VVG 0134), om onder kwarantyn sy virulensie gedurende infeksie van spanspekplante te verander. Saailinge van die spanspekkultivars Early Sweet (geen weerstandsgene), Amber (Fom2) en Fiata (FornI, Fom2) is opeenvolgens in twee sement trêe in 'n gaashuis verbou. Die afsonderlike aanplantings is beëindig sodra gevorderde Fusarium-verwelksimptome verkry is, of nadat vrugte ge-oes is. Vir die eerste aanplanting is oorgeplante Imperial 45 saailinge kunsmatig met die nil mutant isolaat geïnokuleer. Tydens die opeenvolgende aanplantings is saad direk in die besmette grond gesaai ten einde natuurlike infeksie te verkry. Met elke aanplanting is isolasies gedoen vanaf verteenwoordigende plante wat Fusarium-verwelksimptome getoon het. Alle F. oxysporum isolate wat verkry is, is ge-enkelspoor en hul nit mutant status en VVG is bevestig. Virulensie van die gemerkte isolate is bepaal deur inokulasie van die differensiële kultivars. Alle plante van die vatbare Imperial 45 en Early Sweet kultivars wat in trog A geplant is, het Fusarium-verwelksimptome getoon. Die gemerkte isolate wat vanaf die verteenwoordigende plante verkry is, is almal as ras 0 geklassifiseer. Tydens die eerste aanplanting van die weerstandbiedende kultivar, Amber (aanplanting No.5), het 6.7% van die plante Fusarium-verwe1ksimptome ontwikkel. Tydens die tweede en derde aanplanting van Amber het die frekwensie van siektevoorkoms verhoog na 56.6% en 81.8 %, onderskeidelik. In teenstelling met die vatbare kultivars, is slegs ras 2 vanuit die Amber plante met siektesimptome verkry. Soortgelyke resultate is met Imperial 45 en Amber in trog B verkry. Aanplanting van kultivar Fiata het egter 'n dramatiese verlaging in die voorkoms van Fusarium-verwelk bewerkstellig. Tydens die eerste Fiata aanplanting (aanplanting No.6) het 1.2% plante Fusarium-verwelksimptome ontwikkel, en 4% tydens die laaste aanplanting. Vanaf hierdie plante is slegs ras 1,2 isolate verkry. Hierdie bevindings, en die feit dat 'n aansienlike hoeveelheid van die Amber (ongeveer 7-15%) en Fiata plante (ongeveer 2%) siektesimptome getoon het, bewys dat FOM vinnig van virulensie verander wanneer die patogeen 'n weerstanbiedende kultivar infekteer. Die vermoë van RAPD analise om tussen isolate wat in virulensie verander het, te onderskei, is ondersoek. Vier isolate van F. oxysporum f. sp. niveum is as 'n buite-groep ingesluit. Om te bevestig dat die isolate wat van ras verander het wel egte vaskulêre patogene is, is 'n histopatologiese ondersoek gedoen. Die drie inleiers wat gebruik is, het die 12 FOM isolate duidelik van die vier F. oxysporum f. sp. niveum isolate onderskei. Die 12 FOM isolate wat drie fiosologiese rasse (ras 0, ras 2, ras 1,2) verteenwoordig het, is egter saam gegroepeer, wat aandui dat hierdie metode nie tussen rasse van FOM kan onderskei nie. Inokulasiestudies met die differensiële kultivars het die virulensie van die isolate bevestig. Die histopatologiese ondersoek het verder bewys dat beide FOM rasse (ras 2, ras 1,2) wat vanaf die wilde tipe ras ° isolaat ontstaan het, hul vermoë behou het om as egte vaskulêre patogene op te tree.

Thesis (PhD (Botany and Zology))--University of Stellenbosch, 2006. / The well-known, and often phytopathogenic, ophiostomatoid fungi are represented in South Africa by the two phylogenetically distantly related genera Ophiostoma (Ophiostomatales) and Gondwanamyces (Microascales). They are commonly associated with the fruiting structures (infructescences) of serotinous members of the African endemic plant genus Protea. The species O. splendens, O. africanum, O. protearum, G. proteae and G. capensis have been collected from various Protea spp. in South Africa where, like other ophiostomatoid fungi, they are thought to be transported by arthropod vectors. The present study set out to identify the vector organisms of Protea-associated members of mainly Ophiostoma species, using both molecular and direct isolation methods. A polymerase chain reaction (PCR) and taxon specific primers for the two Protea-associated ophiostomatoid genera were developed. Implementation of these newly developed methods revealed the presence of Ophiostoma and Gondwanamyces DNA on three insect species. They included a beetle (Genuchus hottentottus), a bug (Oxycarenus maculates) and a psocopteran species. It was, however, curious that the frequency of these insects that tested positive for ophiostomatoid DNA was very low, despite the fact that ophiostomatoid fungi are known to colonise more than 50% of Protea infructescences. Subsequent direct isolation methods revealed the presence of reproductive propagules of Ophiostoma spp. on four Protea-associated mite species (Oodinychus sp., two Tarsonemus spp. and Proctolaelaps vandenbergi). These mites are numerous within Protea infructescences and Ophiostoma spp. were isolated from a high frequency of these individuals. The Oodinychus sp. mite was found to vector most of the Protea-associated Ophiostoma species. It was thus postulated that the mites (in particular the Oodinychus sp.) act as primary vectors of the Protea-associated Ophiostoma species. The association between Oodinychus mites collected from P. repens and O. splendens proved to be mutualistic. Mites feeding on this fungus showed significantly higher population growth than mites feeding on any of the other fungal species tested. The short- and long-distance dispersal methods of these mites were also investigated. Firstly the ability of mites to move from drying infructescences to moist and sheltered areas such as provided by intact infructescences on the same plant was investigated experimentally. Significantly more mites were found to actively disperse from drying infructescences to artificially manufactured infructescences containing moistened filter paper shreds than to artificially manufactured infructescences containing dry filter paper shreds. The frequent fires associated with the habitat of these mites would, however, require movement over larger areas than what would be possible through self-dispersal. Dispersal of mites via air currents was thus investigated using sticky traps, but no Ophiostoma-vectoring mites were captured in this way. Self-dispersal aided by air currents could thus be ruled out, and our investigations shifted to vectored dispersal. Numerous insects emerging from Ophiostoma-containing P. repens and P. neriifolia infructescences were collected using specially designed emergence cages. Scanning electron microscopy and stereo-microscopy revealed that all three Ophiostoma-vectoring mite genera were phoretic on the beetle G. hottentottus. Tarsonemus spp. and P. vandenbergi were also phoretic on the beetles Trichostetha fascicularis and T. capensis associated with P. repens and P. neriifolia flowers. Mites collected from the surface of these beetles were found to vector reproductive propagules of various Ophiostoma spp. This thus seems to be the only method of long-distance dispersal of these mites and subsequently also the Protea-associated Ophiostoma species. Molecular phylogenetic reconstruction based on large subunit, ITS and beta-tubulin DNA sequence data suggests a polyphyletic origin for the Protea-associated members of Ophiostoma, which proposes multiple invasions of this unusual niche by these fungi. These studies also revealed the presence of four new species of Ophiostoma associated with Protea spp. The new species O. palmiculminatum, O. phasma, O. gemellus and Sporothrix variecibatus were thus described. Ophiostoma palmiculminatum is associated with P. repens infructescences and the Oodinychus mites collected from them. Ophiostoma phasma was collected from various Protea and mite species. Ophiostoma gemellus and Sporothrix variecibatus were initially only isolated from mites, but have subsequently also been isolated from Protea spp. The present study clarifies many aspects pertaining to the phylogeny and ecology of the interesting members of Ophiostoma associated with Protea hosts. As such this study will form the platform for further studies on the co-evolution of these insect / mite / fungi / plant associations.