Many efforts have been made to improve sorghum [Sorghum bicolor (L.) Moench] varieties, but adoption of improved varieties remains low. Sorghum has diverse panicle architecture and grain qualities that vary within and between races, and utilization and adoption may depend on these traits. Recent efforts in West Africa to improve local guinea race germplasm as a base material have diversified potential options: there are breeding materials with a range of panicle types with increased grain number per panicle and a range of droopiness, as well as laxness and threshability. This study was designed to expand our understanding about sorghum grain and panicle traits that are important for farmers in the Sudan savanna zone of Mali. We combined a sorghum panicle sorting activity with qualitative interviews in Mande and Dioïla to understand farmers' knowledge and preferences about sorghum characteristics. A total of 20 panicle sorting activities and 20 interviews were conducted with men and women sorghum producers. Based on their roles and responsibilities in sorghum production and processing, farmers associated specific panicle types, plant types, and grain traits with aspects of pest control, threshability, storage duration, and yield. Farmers preferred open panicles and droopy architecture for disease and pest control; hard grains for storage and appropriate ratios of flour and grits; high density of grain on the panicle for yield; and specific glume qualities for threshability. Breeding programs need to consider these regional preferences and gender roles to develop appropriate material and increase adoption of productive varieties.

Sorghum is a staple food crop in Niger and its production is constrained by sorghum midge and the use of low yielding, local sorghum varieties. To improve sorghum productivity, it is crucial to provide farmers with high yielding sorghum cultivars that are resistant to midge. We evaluated 282 genotypes in four environments of Niger Republic. Alpha (0.1) lattice with two replications was the experimental design. Genotype and genotype by environment (GGE) biplot analysis was used to study grain yield (GY) stability and G × E interactions. The results revealed that two distinct mega environments were present. Genotype L232 was the best genotype for GY in the first planting date at Konni and the first and second planting dates (PDs) at Maradi. Genotype L17 was the best for GY in the second PD at Konni. The second PD at Konni was the most discriminating environment while the first PD at Konni is suitable for selecting widely adapted genotypes for GY.