| dc.description.abstract | This study was conducted to enhance the efficiency of IITA-Uganda cassava breeding programme through quantification of genetic diversity among the cassava mosaic disease (CMD) resistant breeding populations and identification of morphological markers for traits of agronomic interest. Field trials were planted with 200 CMD resistant clones and 7 local landraces in a randomised complete block design with 3 replicates at the IITA Sendusu station in Namulonge (Wakiso district) in the second rains of 2006 (2006B) and the first rains of 2007 (2007A). In both seasons, morphological characterisation was done using a descriptor list and disease attack scored using standard procedures. Storage root yield parameter data were collected at 12 months after planting. In 2006B, one sample of tender young leaves was collected per clone at 3 weeks after planting for molecular characterisation using simple sequence repeat markers at the National Agricultural Research Laboratories Institute, Kawanda. Non-parametric correlations and Chi square tests were used to identify suitable morphological markers for storage root yield and disease susceptibility within the 2006B and 2007A datasets. Principal component analysis and hierarchical clustering were used to determine the genetic diversity from the observed morphological variability. The structure of genetic relatedness was determined from dendograms and multidimensional scaling plots. Allele frequencies, gene frequencies and heterozygosity were computed per locus as indices of genetic diversity from the molecular data. Molecular variability dendogram stability was assessed using Bootstrap values and a cophenetic correlation coefficient. Petiole length, leaf lobe length and width and stem growth habit were identified as suitable morphological markers for storage root yield, and light green apical leaf colour for CBSD susceptibility. The genetic diversity from morphological variability was low, with coefficients of similarity ranging from 0.73 to 0.98. It was also low from molecular variability with mean gene frequency and heterozygozity of 0.24 and 0.28 per locus, respectively. However, molecular variability dendogram Bootstrap values (< 20%) and cophenetic correlation coefficient (0.46) were very low probably due to poor band resolution of agarose gel that was used instead of PAGE. It was recommended that critical values for the identified morphological markers be determined and the genetic base be increased through exotic introductions and introgression of desirable traits through marker assisted selection. Future similar studies should use PAGE instead of agarose gel.
Key words: Cassava improvement, Genetic diversity, Manihot esculenta Crantz, morphological markers, Simple Sequence Repeat markers, Uganda | en_US |
