| dc.description.abstract | Bananas are both a food and cash crop in Uganda’s southwestern district of Ntungamo. The soils have low inherent fertility and low CEC. This is problematic for monovalent cations like exchangeable K, which is also taken up in large amounts by bananas. Potassium is the most limiting nutrient for banana production in Uganda. Application of K to address its deficiency requires knowledge of its critical value in the soil, which hitherto is ambiguous. Soil fertility reduces with distance from the homestead, but it is not clear how strongly this affects the K application rates for improving (sufficiency rate) and sustaining (maintenance rate) banana productivity. The specific objectives of this study were therefore to determine the critical value for exchangeable K, and to evaluate the variability of K sufficiency, maintenance and total application rates with distance class from the homestead in Ntungamo district. The study was carried out in Butare village in Nyakyera subcounty, Ntungamo district on ten farms. Fisteen mats of the banana cultivar ‘Enyeru’ were selected on each of the farms from the distance classes ‘near’ (<20 m from homestead), ‘mid distance’ (20-39.9 m from homestead) and ‘far’ (≤40 m from homestead). The fresh bunch weight at horticultural maturity was taken using a salter scale and yield (t ha-1 yr-1) computed for each reference mat. Total number of fingers per bunch were counted and recorded. Five soil cores were randomly taken from 0-30 cm depth layer around each reference mat, composited and taken to the laboratory for analysis. Exchangeable K was extracted from each sample using Mehlich 3 solution and read using an atomic absorption spectrophotometer. A scatterplot of fresh bunch yield vs. exchangeable K was subjected to boundary line analysis to identify points representing the attainable yield for a given level of exchangeable K. The functional relationship between attainable yield and exchaneble K was determined using linear optimsation conditioned on minimising the Root Mean Square Error (RMSE) of the oberved vs. predicted attainable yield, and the associated squared correlation coefficient (r2) evaluated. From the fitted function, the critical value of exchangeable K for banana production in the study area was computed as the value of independent variable required for 90% of the maximum yield (60 t ha-1 yr-1) observed in the study area. The reference mats whose soil K test value was less than the critical value had a sufficiency rate computed. And the result divided by 5 to obtain an annual application rate (kg K ha-1 yr-1) for gradually raising the soil K level to the critical value. The K maintenance rate (kg K ha-1 yr-1) was computed from the annual K exports in harvested bunches. The sum of the K maintenance rate and annual K sufficiency rate gave the total K application rate needed to increase and sustain banana productivity for a given reference yield. Two-way analysis of variance was used to assess the variation in K sufficiency rate, K maintenance rate and total K application rate across distance classes from the homestead. There was a strong functional relationship between attainable fresh bunch yield and exchangeable K (RMSE = 0.1; r2 = 0.93). The critical value of K was found to be 0.53354 cmolc kg-1. The sufficiency K rate for banana production increased significantly from ‘far’ than at the near distance class from the homestead. However, there was no significant difference in the maintenance K rate across the distance classes. The total K application rates was about 120 kg K ha-1 yr-1. It is recommended that this study be conducted across all the major banana agroecological zones of Uganda and also to find suitable nutrient input materials for meeting the total K requirement for banana production, which farmer can afford. | en_US |
