Serba PearlMillet Hays

Assessment of Pearl Millet Production Problems in West Africa and Molecular Diversity Analysis of Pearl Millet Parental Lines

Award Amount

$29,904 (2013 - 2019)

Research Team

Desalegn Serba        Kebede Muleta
Ramasamy Perumal

U.S.A. collaborating institutions

Kansas State University - Western Kansas Agricultural Research Center in Hays

International collaborating institutions

Burkina Faso - Institut de l'Environnement et de Recherches Agricole (INERA)
Mali - Institut d'Economie Rurale (IER)
Niger - International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Institut National de la Recherche Agronomique du Niger (INRAN)
Senegal - Centre d’Etudes Régional pour l’Amélioration de l’Adaptation à la Sécheresse (CERAAS), Centre National de Recherche Agronomique (CNRA), Institut Sénégalais de Recherches Agricoles (ISRA)


Pearl millet is a warm-season annual cereal crop grown for food in the hottest and driest tropical regions of Africa and Asia. It is the most drought tolerant of all domesticated cereals and a staple food crop for about 90 million people across the Sahel of Africa and fringes of the Thar desert in India, where severe abiotic stress limits the growth of other crops. It is grown where no other cereal will yield grain or is likely to fail due to heat, drought, and poor soil fertility. Pearl millet is also grown in the United States in limited areas as a forage and cover crop. It has also a high potential to be a commercial grain crop in drought-prone areas due to its water and nutrient use efficiency. The grain is nutritious, has higher protein and energy levels than corn or sorghum, and has been evaluated positively in feed rations of chickens and steers. The productivity of pearl millet is still challenging in drought-prone areas.




The main goal of the pearl millet breeding program at the Kansas State University Agricultural Research Center in Hays, Kansas is to develop parental lines for producing hybrids for grain and forage use. The research project focuses on harnessing genetic variability and the development of parental lines for new hybrids with high genetic yield potential, yield stability, improved drought and high-temperature tolerance, disease resistance, and enhanced nutritional quality especially micronutrient density of the grain. This project addresses the production constraints in the drought-prone areas of western Kansas and in major producing West African countries: Senegal, Niger, Mali, and Burkina Faso.



This project and research was funded by the Feed the Future Innovation Lab for Collaborative Research on Sorghum and Millet, known as the Sorghum and Millet Innovation Lab (SMIL). This lab is funded by USAID and managed at Kansas State University.

Project Description

Pearl millet is an important staple food and fodder crop in West Africa especially in Mali, Senegal, Niger, and Burkina Faso. However, the productivity is still low as compared to the genetic potential of the crop. Different biotic and abiotic constraints are expected to contribute to the low productivity. To identify a priority area for future research a professional assessment was conducted through informal survey and preliminary evaluation nursery of germplasm. A total of 100 entries comprising of inbred lines, experimental hybrids, and open pollinated germplasm were assembled and evaluated in Niger and Senegal. Informal survey of the production problems were conducted in all four countries to identify a boarder-cutting production problem that needs a collaborative research intervention.

Pearl millet breeding research at the Kansas State University Agricultural Research Center in Hays, Kansas developed several parental lines in the 1980s and 1990s using mainly phenotypic evaluation. The breeding program also assembled germplasm from various sources. Molecular diversity analysis of these materials identified novel alleles for different important traits. Therefore, a next-generation sequencing technology called genotyping-by-sequencing was used to genotype the materials and diversity analysis conducted using high throughput SNP markers. The outcome of this diversity analysis has helped in founding preliminary heterotic groups and conducting an efficient hybrid breeding program.

Principal Investigator

Dr. Desalegn Serba

Kansas State University
Research Geneticist
USDA Agricultural Research Service
US Arid-Land Agricultural Research Center
Phone: 520-316-6368
Email: [email protected]

Regional Impact

Burkina Faso - Ouahigouya, Ougadougou
Mali - Segou, Koutiala
Niger - Aguié, Kollo, Boboye
Senegal - Bambey, Nioro du Rip

Country Coordinators

Dr. Moustapha Moussa

Institut National de la Recherche Agronomique du Niger (INRAN)

Dr. Ndjido Kane
Director & Coordinator

Institut Sénégalais de Recherches Agricoles du Senegal (ISRA)

Progress and Impact

The accessions, hybrids, and inbred lines evaluated at two representative locations in the Sahelian zone produced several high-yielding hybrids and accessions with good level of downy mildew resistance were identified. A total of 400 samples were sequenced three times on Ion Proton. More than 540 million unique reads were obtained. A total of 103,186,800 markers were identified using the TASSEL5-reference pipeline against the pearl millet reference genome. After filtering the SNPs for 20% missing, 1% MAF, and InDels 82, a total of 112 genome-wide SNPs markers were discovered. An approximate average marker density of 48.3 per Mb of the genome was calculated.

The population structure analysis identified six sub-populations that mostly overlap with the genetic origin of the germplasm accessions/source of inbred lines. The principal component analysis showed the population structure and within subpopulation diversity as PCA1 and PCA2 explained less than 10% of the genetic diversity of the population. A neighbor-joining phylogeny analysis grouped the materials into 12 sub-clusters with West Africa as a base. Most clusters were formed among the inbred lines developed in the U.S. Genome-wide linkage disequilibrium analysis and findings show the faster LD decay in West African sub-population than the rest of the sub-populations. Assessment of genome-wide patterns of nucleotide variation within each sub-population revealed that average genome-wide FST is higher in the Middle East followed by East Africa. The differentiation within Indian sub-population was low.

Presentations and Publications