COMPARATIVE STUDY ON MULTIPLE RESPONSE OPTIMIZATION USING RESPONSE SURFACE METHODOLOGY TECHNIQUE- CASE OF FRENCH BEANS AT KARIUA SUBLOCATION IN KANDARA, MURANG’A COUNTY.

Abstract

In Kariua sub-location, the resources like land and man-power are extremely scarce and this has affected French beans production negatively. This research, conducted at Kariua, aimed to optimize and compare French beans output and plant health simultaneously using manures, water and crop-spacing as control parameters. The response variables of interest were average pods’ mass, infected leaves and unharvested pods. Manures and seeds were obtained from the farmers while water was drawn from river Irera. A survey was conducted at the beginning of the experiments to assess the current situation in French beans production and to provide the factors’ levels at which the experiments were to be conducted. All responses and control factors were measured per crop point. The survey results showed that farmers were experiencing low yields on average and poor plant health (harvest=13.4 g, infected leaves= 8 and immature pods= 15) which were attributed to scarce resources, pests, infections and poor farming techniques. Soil testing analysis was carried out and experiments were performed along the lines of untested and tested soils using response surface methodology and Hoke D2 design. For the untested and tested soils respectively, the diammonium phosphate and calcium ammonium nitrate applications were 5.65 dg and 2.65 dg, and 2.5 dg and 2.5 dg. Only one variety of the beans, Gregor, was tested since it was the variety cultivated that time. For both cases, 2nd order models were fitted and the two sets of models were compared using optimized responses’ results when optimal levels of factors were applied in replicates. The optimizing factor levels for manure, water and spacing were found to be 26.1 g, 4.0 ℓ and 11.1 cm, and 24.5 g, 4.1 ℓ and 10.3 cm for untested and tested soils’ cases respectively. The theoretical and practical optimal responses were found to be in agreement. The practical optimal responses on average were 24.2 g, 0 leaves and 21 pods, and 28.5 g, 0 leaves and 32 pods for untested and tested soils’ cases respectively. These results were found to be statistically far much better than what the farmers are currently experiencing while those from tested soils’ case were the best because of larger mean values and p-values less than 0.05 level of significance compared to the other cases. Therefore, Kariua farmers can apply the optimal factor levels for the tested soils’ case for maximum benefit from the very limited resources in the region.