Spatial Multivariate Cluster Analysis for Defining Target Population of Environments in West Africa for Yam Breeding

Abstract

Yam (Dioscorea spp.) is a major staple crop with high agricultural and cultural significance for over 300 million people in West Africa. Despite its importance, productivity is miserably low. A better understanding of the environmental context in the region is essential to unlock the crop's potential for food security and wealth creation. The article aims to characterize the production environments into homologous mega-environments, having operational significance for breeding research. Principal component analysis (PCA) was performed separately on environmental data related to climate, soil, topography, and vegetation. Significant PCA layers were used in spatial multivariate cluster analysis. Seven clusters were identified for West Africa; four were country-specific; the rest were region-wide in extent. Clustering results are valuable inputs to optimize yam varietal selection and testing within and across the countries in West Africa. The impact of breeding research on poverty reduction and problems of market accessibility in yam production zones were highlighted.