Thesis (MBA)--Stellenbosch University, 2007. / ENGLISH ABSTRACT: The importance of improving the net farm income of production units are an economical reality. One of the most important factors determining the profitability of deciduous fruit like apples and pears is the difference between the market price and the production of different cultivars. The profitability of production units is therefore determined by the cultivar composition. A decision support system is proposed for production units in the deciduous fruit industry to reach the optimal cultivar composition and an implementation plan of how to reach this optimal composition. The purpose of this study will include the optimising and adjustment of different cultivars to reach the highest possible profitability within the ranges of the available resources. The study will also contain all the relevant facts and information which are needed to make a choice between the different cultivars. The decision making process of deciduous fruit producers could be assisted with a decision support system or model. The use of such a decision support system or model to reach the optimal cultivar composition could lead to economic advantages for deciduous fruit producers. A wide range of choices regarding certain procedures and methods are found in the decision making process. Modelling and simulation are some of these methods that are used in the agricultural industry to assist the industry on a micro and macro level. Linear programming is one of these modelling methods which can be used to determine the optimal cultivar choice. Linear programming involves the creating and solving of optimisation problems with linear objective functions and linear constraints. linear programming is a very powerful tool that can be applied in many business situations. The optimal cultivar choice is only theoretical if this composition could not be simulated in a model over a certain period of time to detennine the economical feasibility of the proposed cultivar composition. The end results of such an optimisation problem provide to the user of this linear programming model a powerful tool to assist producers in its decisions regarding cultivar choices. The average result of a 90% increase in profitability over the group for whom the study was conducted emphasized the support that this model can give to deciduous fruit producers in making the correct cultivar choice decisions / AFRIKAANSE OPSOMMING: Dit is 'n ekonomiese realiteit vir enige produksie-eenheid om sy netta boerdery inkomste so haag as moontlik per hektaar te kry. Die grootste faktor wat die winsgewendheid van sagtevrugte soos byvoorbeeld appels en pere beinvloed, is die verskil in markpryse en produksie tussen die verskillende kultivars. Die kultivarsamestelling van produksieeenhede bepaal die winsgewendheid van so 'n eenheid. Om oplossings vir optimale kultivarwysigingstrategiee te bekom word 'n besluitnemingsondersteuningstelsel vir die sagte vrugte produksie-eenhede voorgestel. Die doel van die studie is vir die optimalisering van verskillende kultivars, of kultivarwysigings, om wins te maksimeer binne die grense van beskikbare infrastruktuur. Die studie sal alle relevante inligting en feite bevat wat in ag geneem moet word wanneer die keuse tussen kultivars gemaak word. Die besluitnemingsproses van sagtevrugte produsente kan ondersteun word met 'n besluitnemingsondersteuningstelsel of -model. Die gebruik van 'n besluitnemingsondersteuningstelsel of -model, en keuses tussen kultivars vir toekomstige aanplantings te maak, kan tot ekonomiese voordele vir die produsente lei. 'n Wye reeks van keuses in terme van benaderings, prosedures en metodes word in die besluitnemingsproses gevind. Modellering en simulasie vorm deel van hierdie metodes wat help am verskillende landbou industriee of sisteme op 'n mikro sowel as makro vlak te verstaan. Lineere programmering is een van hierdie modellering metodes wat gebruik kan word om 'n optimale samesteliing van kultivars te verkry. Uneare programmering is 'n optimaliseringstegniek waarin beide die doelwit en aile beperking5 as 'n reeks van lineere vergelykings uitgedruk word. Uneare programmering spreek dus die optimaliseringsprobleem binne sekere beperkinge direk aan en word daarom universeel as waarskynlik die algemeenste en wydste optimaliseringstegniek gebruik. Die optima Ie kultivarsamestelling is slegs teoreties van aard indien hierdie kultivarsamestelling nie in 'n model gesimuleer kan word oor 'n sekere tydperk, sodat die ekonomiese haalbaarheid van so 'n verandering geregverdig kan word nie. Die eindresultaat van hierdie optmaliseringsprobleem bied aan die gebruiker van hierdie lineere programmeringsmodel 'n kragtige instrument wat die produsent kan ondersteun met sy besluitneming rakende die optimale kultivarsamestelling . Die navorsingstudie se groepsgemiddelde resultaat van 90% winsverbetering, binne die raamwerk van beperkinge wat gebruik is, bied voldoende ondersteuning vir die effektiwiteit van die optimaliseringsmodel wat ontwikkel is.