Crop-Environment-Management Interaction Models for Predicting Yield Potential and Yield Gaps in Nigerian Smallholder Systems

This study develops and validates crop-environment-management interaction models for predicting yield potential and quantifying yield gaps in Nigerian smallholder farming systems, contributing an original modelling framework for evidence-based production frontier analysis. Yield gap analysis, comparing actual farm yields to attainable potential yields, is a powerful tool for prioritising agronomic research and extension investment. However, accurate yield gap quantification in Nigeria requires models that are specifically calibrated to local crop varieties, soil types, and farming conditions rather than parameterised from non-African data sources. This study calibrates and validates the APSIM and DSSAT crop simulation platforms for major Nigerian crops including maize, sorghum, rice, and cowpea using extensive variety characterisation data, multi-year multi-site experimental datasets from IITA and NAERLS trials, and soil characterisation data from 120 sites across six agro-ecological zones. Simulated potential yields under water-limited and nutrient-limited scenarios are compared with farmer actual yields from the LSMS-ISA Nigeria panel survey. Yield gap decomposition analysis attributes gap components to water, nutrient, and management deficiencies. Findings reveal national average yield gaps of 74 percent for maize, 68 percent for rice, and 58 percent for sorghum between actual and attainable yields. Nutrient management deficiencies account for the largest share of yield gaps in the Guinea savanna and forest zones, while water deficiency dominates in the Sudan-Sahel zones. Model outputs identify the most impactful single management change for gap closure in each environment. The study contributes calibrated APSIM and DSSAT models for Nigeria and recommends a national yield gap monitoring programme.

Keywords: yield gap analysis, crop simulation models, APSIM, DSSAT, Nigerian smallholder.