In order to assess the performance of water management approaches and irrigation systems used by the sugar industry in the Lowveld of Zimbabwe, a sugarcane yield and irrigation systems simulation model was developed. The model, named ZIMsched 2.0, was used to predict how field derived indices of irrigation systems performance, such as the coefficient of uniformity, CV, impacted on estimated recoverable crystal, ERC, yields and the water balance. This was done across a range of soil conditions, seasonal climates, irrigation system types and existing and refined irrigation scheduling strategies. Results of a verification study of the model showed an index of agreement, 'd', equal to 0.96 and a Pearson's correlation coefficient equal to 0.94, between observed and simulated yields of ERC, relative to a reference treatment. Application of the model showed the actual and also the potential performance of the different irrigation system hardware. Additional applications of the tools and information which were developed as a result of this research included an integrated economic assessment of peak irrigation system design specifications and associated deficit irrigation watering strategies. In an effort to translate theoretical water savings into practical realities a range of novel water management tools was also developed. Most of the drip irrigation systems in the Lowveld were performing below potential due to excessive infield variations in applied water. The performance of furrow irrigation systems was limited by the large variations in water applied to individual furrows, and water applications that were, on average, excessively high relative to soil water holding characteristics. Simulations showed that sub-surface drip irrigation systems have a slight edge on other irrigation systems in terms of potential efficiency. Average water savings for drip irrigation systems ranged from approximately 2.2 to 1.5 Ml/ha relative to floppy irrigation systems, and 3.5 to 2.3 Ml/ha relative to typical furrow irrigation systems, depending on how water applications were scheduled. A major finding was that there was potential for the Lowveld sugar industry to use up to 30% less water per hectare on an annual basis if ZIMsched, a specialist spreadsheet-based irrigation scheduling tool developed during the course of the project, was used to derive more appropriate and system specific water management guidelines. However, simulations showed that with the more precise irrigation scheduling there could be a slight crop yield penalty when the distribution uniformity of applied water was poor.

Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2004.