Clustering of the Chemical Properties of Small‐scale Farm Soils to Develop Fertilizer Blend Ratios

Crop management through site-specific fertility methods is crucial for both economic and environmental benefits. Potato-specific fertilizer has not been fully established in Kenya, and the current fertilizer recommendation has been in place for a long time. A study was designed through soil sampling from small-scale farms to make a soil-based fertilizer recommendation for potato production. Inverse distance weighting interpolation was used to study spatial distribution of pH and major nutrients. Cluster analysis was used to group sampled farms into similar fertility clusters for easier fertilizer recommendations. Spatial distribution maps were identified for the project area. Three and four clusters were established in Meru and Nyandarua counties, respectively, among which three had adequate nutrients for potato production. At least 150 kg N ha-1 is recommended in most of the farms and 160 kg P ha-1 in Nyandarua, while 285 kg K ha-1 was recommended in one of the clusters (nitrogen, phosporus, potassium [NPK] 1:1:1.5). Cluster analysis can be used to enhance fertilizer recommendations. The assumption that K is adequate in tropical soil should not be generalized. Fertilizers in the market with formulations close to the recommended ones should be evaluated for field performances and adjustment of the final potato fertilizer. Inverse distance weighing interpolation clearly showed spatial distribution of pH and major soil chemical properties for better fertility management. Cluster analysis of soil chemical properties was able to delineate farms that could have the same fertilizer recommendation. There is a need for different fertilizer recommendations between Meru and Nyandarua.