OKWU, QUEEN UDODIRIM
BREEDING FOR IMPROVED STARCH QUALITY TRAITS IN CASSAVA (Manihot esculenta Crantz)
ABSTRACT
Knowledge on the genetic diversity of cassava for quality traits is a pre-requisite to develop varieties that meet the need of industries in terms of starch traits. However, progress in improving starch quality traits has been limited mainly because current methods for evaluating them are laborious, expensive and prone to errors when screening large number of clones. The research objectives were aimed at developing calibration models for starch quality traits using a high-throughput phenotyping tool, near infra-red spectroscopy (NIRS) and identifying genomic regions and candidate genes associated with starch quality in cassava as well as selecting for genotypes that are stable for these traits across the test environments. Calibration models were developed using partial least square (PLS) regression. The results of the statistical modelling indicate that robust calibrations were established for some traits especially for dry matter content (Rc= 0.84, RCV = 0.83 and SECV = 0.47); starch yield (Rc = 0.81, RCV = 0.76 and SECV = 0.54) and amylose content (Rc = 0.74, RCV = 0.74 and SECV= 0.23) indicating that NIRS was reasonably accurate for the screening of these traits.The Genome-wide Association study (GWAS), was used to identify the presence of genes that could influence starch quality in cassava. Using mixed linear models, GWAS detected significant SNP markers associated with all the starch quality traits studied. The presence of significant SNP markers identified to be associated with more than one trait is likely an indication of pleiotropic effects or closely linked genes that regulates these traits. GWAS also detected seven candidate genes belonging to the glycosyl transferase, glycoside hydrolase family1 and sugar transporter families that were linked to the starch quality traits in cassava. These identified genes could represent new sources of alleles for breeding efforts to develop improved cassava starch cultivars. The stability of genotypes for the starch quality traits in cassava were assessed using Genotype main effect and genotype-by-environment interaction (GGE) biplot model and multi-trait index. Analysis of variance (ANOVA) revealed significant genetic variations for most traits indicating opportunity for selection for the improvement of these traits in cassava. ANOVA also revealed significant environmental effects for most traits, thus, justifying the need for Genotype x Environment interaction study to identify genotypes that performed best for the quality traits studied across the test environments. GGE biplot analyses showed that the genotype NR17C2aF60P017 had the highest mean for peak, final and setback viscosities and it was stable across the test environments while the genotype NR17C2aF60P014 was stable across the test environments with the highest means for dry matter content and starch yield. The multi-trait index using FAI-BLUP index revealed 42 genotypes ranked based on high mean performance and stability for all the cassava starch quality traits studied. The genotype NR17C2aF60P014 ranked topmost in performance and stability for all the traits. This study provides critical information for fast-tracking the improvement of starch quality in cassava.