Thesis (MComm)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: Decision making in a complex environment is not an easy step as there are uncertainty variables that cannot be foreseen. This causes decision makers to use tools that can help and support difficult decision making in such a complex environment. These tools can be different variations of which a model that imitates the environment or system are the most commonly used. The systems approach is normally used to describe a complex environment. Such an environment consists of elements that are linked together to reach a goal or perform a function. A model is thus used to simplify the reality and to imitate and simulate the system as close as possible. Models are built in different forms of which mathematical and physical models are the main types. Physical models imitates the system through physical measures whereas mathematical models make use of equations that are interdependent. The probability that a single simulation of a mathematical model will represent reality is very rare. To overcome this problem a stochastic approach can be followed where a series of possible outcomes can be simulated for a set of variables. Hereby a probability distribution can be generated for a specific outcome. In this study a stochastic simulation model is used as a decision support tool for table grape producers where the impact of different scenario’s can be evaluated. The model is developed from an existing model for long term crops and adjusted for table grape production. Table grape producers and policy makers can use the model for decision making and scenario planning. / AFRIKAANSE OPSOMMING: Besluitneming in ‘n komplekse omgewing is gewoontlik nie ‘n eenvoudige stap nie aangesien daar veranderlikes is waarvoor nie voorsien kan word nie. Dit bring mee dat besluitnemers na hulpmiddels soek om moeilike besluite in so ‘n komplekse omgewing te ondersteun. Hulpmiddels kan verskeie vorme aanneem waarvan ‘n model wat die omgewing of stelsel naboots die mees algemene metode is. Die stelselsbenadering word in die algemeen gebruik om ‘n komplekse omgewing voor te stel. So ‘n omgewing bestaan normaalweg uit verskeie elemente wat aan mekaar gekoppel is om ‘n doel te bereik of ‘n funksie te verrig. ‘n Model word dus gebruik om die werklikheid te vereenvoudig en om die stelsel van belang so na as moontlik na te boots of te simuleer. Modelle kan verskeie vorme aanneem met wiskundige en fisiese modelle wat van die hoof tipes is. Fisiese modelle poog om die stelsel na te boots met fisiese maatstawwe, terwyl wiskundige modelle die stelsel naboots deur wiskundige vergelykings wat interafhanklikheid in ag neem. Die waarskynlikheid dat ‘n enkele simulasie van ‘n wiskundige model die werklikheid sal verteenwoordig is laag. Om die probleem te oorkom kan ‘n stochastiese benadering gevolg word waar ‘n reeks van uitkomste gegenereer word vir ‘n stel veranderlikes. Sodoende kan ‘n waarskynlikheidsverdeling gegenereer word vir spesifieke uitkomste. In hierdie studie is gebruik gemaak van ‘n stochastiese simulasiemodel om as besluitnemingshulpmiddel te dien vir tafeldruifprodusente waar die invloed bepaal kan word wat verskeie scenario’s op die prestasieparameters het. Die model is opgestel vanaf ‘n bestaande model vir langtermyngewasse en aangepas vir tafeldruifproduksie. Hierdeur kan tafeldruifprodusente en beleidmakers gebruik maak van die model om as hulpmiddel te dien tydens beoogde besluitneming en scenariobeplanning.