Decision support system for site-specific fertilizer recommendations in cassava production in Southern Togo

Abstract

The Quantitative Evaluation of the Fertility of Tropical Soils (QUEFTS) model recommended as a decision support tool for deriving optimal site-specific fertilizer rates for cassava has limited ability to estimate water-limited yields. We assessed potential and water-limited yields based on the light interception and utilization (LINTUL) modelling approach in order to enhance the determination of fertilizer requirements for cassava production in Southern Togo. Data collected in 2 years field experiments in Sevekpota and Djakakope were used. Potential ranged from 12.2 to 17.6 Mg ha-1, and water-limited yields from 10.4 to 14.5 Mg ha-1. The simulated average fertilizer requirements were 121 kg N, 2 kg P and no K ha-1 for a target yield of 9.3 Mg ha-1 at Sevekpota, and 103 kg N, 6 kg P and 175 kg K ha-1 for a target yield of 9.7 Mg ha-1 at Djakakope. The variability of fertilizer requirements was attributed to differences in indigenous soil fertility and water-limited yields. The latter correlated well with rainfall variability over years and sites. Integrating LINTUL output with QUEFTS helped account for location- specific weather seasonal variability and enhanced assessment of fertilizer requirement for cassava production in Southern Togo.