An integrative approach based on crop modeling and geospatial and statistical analysis to quantify and explain the maize (Zea mays) yield gap in Ghana

Abstract

In Ghana, maize (Zea mays) is a crop crucial to achieving food and nutrition security. Maize consumption has increased exponentially over the past decades and contributes to 25% of the caloric consumption in the country. In order to assist in decision-making and guide investment in sustainable intensification of maize production, this study set out to identify the determinants of yield and to arrive at potential interventions for closing the maize yield gap. These were quantified using analytical approaches that combine a light use efficiency crop model (LINTUL-1) with statistical and geospatial analyses. Legacy data, auxiliary covariables, and maize fertilizer trials on eight experimental stations in Ghana were used in this study. Overall, the maize yield gap across the stations and trial treatments ranged from 17% to 98%. The variation in yield gap within a single station indicates a significant scope for closing the yield gap through site-specific nutrient management. Multiple linear regression models that explained 81% of the variability in maize yield gap identified soil organic matter, soil water-holding capacity, root zone depth, rainfall, sulfur fertilizer, and nitrogen fertilizer, in that order of importance, as the major determinants for closing the yield gap in the major agroecological zones of Ghana. The yield gap decreased by 1.4 t ha-1 with a 1% increase in soil organic matter. A 1 mm increase of the soil water-holding capacity reduced the yield gap by 1.06 t ha-1, while an increase in pH and in the application of potassium fertilizer widened the gap. These results suggest that both soil physical and chemical properties, together with weather data, should be taken into consideration to arrive at site-specific fertilizer recommendation and other agronomic practices.