Biological pre- and postharvest disease control strategies depend on successful colonisation and survival of the introduced antagonists on the infection court. Effective and economical applications involve targeting the antagonists where they are critically needed i.e. the infection court. Honeybee dispersal is one method of antagonist application to such specific sites. In view of this, an in vitro experiment was conducted to investigate attachment, colonisation and survival of Bacillus subtilis on avocado flowers. Scanning electron microscopy studies showed that the bacterium could attach and colonise avocado flower surfaces. It can also survive on the flowers for longer periods of time. In vivo mode of action of the antagonist against stem-end rot (SER) pathogens was also studied where results showed lysis and degradation of hyphae and conidia. However, no viable colonies were retrieved from bee antagonist dispersal under field conditions. Bee antagonist dissemination was compared with antagonist and fungicide spray applications in terms of SER control and the added effect on other diseases such as Cercospora spot and anthracnose. Spray applications of the antagonist were more effective in reducing the incidence of SER than bee dissemination. Integrated sprays of the antagonist and fungicides significantly reduced the incidence of both pre- and postharvest diseases. The identity of Dothiorella aromatica, one of the most important SER pathogens, was investigated at a molecular level. RAPD techniques using the discriminatory OPC02 primer successfully separated isolates into three groups based on banding profiles. A further study using RFLP identified the pathogen as a Botryosphaeria spp. The most dominant specie was B. parva followed by B. rhodina. Further studies should focus on assessing the distribution of these pathogens within avocado-growing regions of South Africa. / Dissertation (MSc (Plant Pathology))--University of Pretoria, 2006. / Microbiology and Plant Pathology / unrestricted