Validation of the Zambia semi-quantified land evaluation system in Magobbo settlement scheme in Mazabuka District of Southern Province of Zambia

Abstract

A study was conducted in Magobbo area of Mazabuka District to validate the yield predicting capacity of the Zambia Semi-quantified Land Evaluation System for rain fed maize production under small-holder farming conditions. An area where soils had been surveyed and mapped was used as a test site. Areas with three soil types or land mapping units were demarcated and farmers growing maize under rain fed conditions following recommendations by the Farmer Input Support Programme (FISP) were selected on each unit. Using data on the soils and climatic conditions of the area, the expected yields of maize for each land mapping unit were calculated following the methods outlined in the Zambia Semi Quantified Land Evaluation System. The predicted yields were compared with the average yields obtained by farmers in each land mapping unit using a t-test at 0.05 level of significance.When t-test results indicated a non significant difference between the predicted and actual yields, this was considered to be a correct prediction by the system. Results of this study that when yields were predicted using a reference potential yield of 5000 kg/ha, the system made the 33 % correct predictions. Upon correcting the potential yield from the assumed 5000 kg/ha given in the system to the actual potential yield of the maize variety grown of 7000 kg/ha, the yield predicting capacity of the system increased from 33.3 to 66.7 %. A further refinement of the input into the system by taking into account the actual management levels by farmers in land mapping unit 1, which had the largest number of farmers, the predicting capacity of the system decreased to 0% at 5000 kg/ha while that of 7000 kg/ha increased to 75 % which is good. Results of this study have demonstrated that the Zambia Semi -Quantified Land Evaluation System is able to make good predictions of maize yields for small scale farmers, provided correct input parameters such as the potential yield of the crop and accurate assessments of the management levels are used in the system when making yield predictions. There is need to update the potential crop yields of crops used as input in the system for predicting actual yields since the potential yields of a number of crops have significantly changed from what they were in 1987 when the system was developed.