GENETIC ANALYSES AND SELECTION STRATEGIES FOR IMPROVEMEMENT OF FARMER PREFERRED TRAITS IN CASSAVA
Cassava (Manihot esculenta Crantz) is a unique tuberous crop, income earner and highly valued carbohydrate source for close to a billion people especially in Africa. As food demands on the African continent are projected to double in the near future, biotic stresses such as cassava brown streak disease (CBSD) continue to render the most economic part unmarketable. Under high disease pressure, there is a heavy but consistent pattern of damage on the leaves, stem and roots which results in huge yield penalties at the household level. In response, cassava farmers from eleven farming communities were engaged to realign available scientific approaches with farmer preferences and challenges faced. Insights into participatory cassava variety selection as a guide to adoption were revealed. The factors responsible for CBSD resistance in cassava were elucidated in a marker trait association study at the genome level. The relationship between genomic selection and conventional phenotypic selection, as well as their influence on selection response, were examined. Most farmers grow improved varieties but they are subjected to increased CBSD epidemics and varieties with new need to urgently identify new sources of resistance. Cassava production systems under some communities were greatly affected and their income sources heavily fractured after the re-emergence of CBSD. Field evaluations of twelve cassava varieties identified two promising varieties; UG14_2258 and UG14_2391 which met farmers’ preferences. Their performance was better than the most recently released clone, UG11_0017 which was used as a check. Large effect Quantitative trait loci (QTL) on two chromosomes, 4 and 11, conditioning resistance to CBSD were identified. The segment on chromosome 4 co-localized with an introgression segment from the wild relative of cassava, Manihot glaziovii. Strong evidence for interspecific hybridization in the Uganda has been identified. The other identified genome segment was found on chromosome 11 which contains a cluster of nucleotide-binding site-leucine-rich repeat (NBS-LRR) genes. Four candidate genes; Manes.11G131100, Manes.11G130500, Manes.11G130200 and Manes.11G130000, were identified. Further fine mapping studies for the identified candidate genes is essential. The identified genome segments could be useful in guiding marker assisted selection through SNP chip development. This study revealed insights of how useful genome-wide association mapping studies (GWAS) could be in improving prediction accuracies under genomic selection (GS). Response to selection was higher under phenotypic selection compared to genomic selection. Phenotypic selection was more effective than genomic selection, especially for root yield related traits. Both phenotypic and genomic selection approaches had better predictions for disease related traits. Overall, the results suggest that a continuous engagement with farmers to appreciate their preferences will most likely result in active farmer – researcher consortia. This will ably harness existing scientific tools to overcome emerging biotic and abiotic stresses for higher yields.