Genetic diversity of cassava landraces and documentation of farmer's knowledge in Lamu, Kenya

To develop a strategy for improvement of cassava (Manihot esculenta Crantz) productivity for climate change preparedness in Lamu County, Kenya, it was important to understand which cultivars farmers are growing. Landraces should be collected and conserved prior to replacement by improved cultivars to avoid erosion of genetic resources and associated farmer knowledge. In this study, 58 samples of landraces were collected from five administrative wards in Lamu County. These were SNP genotyped using low-density Diversity Array Technologies (DArT) DaRTSeq (TM) together with a reference set of 48 breeding lines, eight coastal and 45 inland landraces to aid identification of the collected samples. Genotyping results defined nine clades of duplicates and six unique clones. The most representative clone was selected from each clade of duplicates and combined with the unique clones giving a set of 15 different genotypes. None of the landraces or improved cultivars used in the reference panel were found in the genotyped collection. Discriminant Analysis of Principal Components revealed four clusters; Lamu germplasm, breeding lines, inland landraces and coastal landraces with low levels of differentiation between them. According to Wrights' F Statistics landraces from Lamu were most closely related to other coastal landraces (Fst = 0.0300), with greater differentiation from breeding lines (Fst = 0.0578) and inland landraces (Fst = 0.0632). Expected heterozygosity was highest for the breeding lines (0.275) and lowest for Lamu landraces (0.240), indicating a narrow genetic base. The results provide valuable information on genetic diversity and identify germplasm for conservation and future use.