Genetic mapping of QTLs affecting productivity and plant architecture in a full-sib cross from non-inbred parents in Cassava (Manihot esculenta Crantz).

Abstract

An attempt was made to identify quantitative trait loci (QTLs) for several productivity and plant architecture traits in a full-sib progeny of 144 individuals from two non-inbred parents in cassava. A molecular linkage map of this cross constructed previously with over 250 markers was the source of molecular markers. The progeny were grown under field conditions at two locations (Palmira and Quilichao) in Colombia and evaluated in 2 years (1998 and 1999) for architecture and productivity traits. Architecture traits evaluated were plant height (PH), branching height (BH), branching levels (BL), branching index (BI), stem portion with leaves (SPL) and leaf area index (LAI). Productivity traits were those related to total dry matter production and distribution, namely fresh root yield (FRY), fresh shoot yield (FSY), harvest index (HI) and the number of storage roots (NR). Phenotypic evaluation of the traits in this population revealed continuous variation for all traits. Broad-sense heritability estimates, ranged from 36% (for NR) to 94% (for BH). Several significant phenotypic correlations were observed between architecture and productivity traits. Primary QTLs, using the single-QTL model, and secondary QTLs, by a primary QTL interaction model, were detected by interval mapping. A total of 30 primary QTLs and 84 secondary QTLs were detected. We identified 35% of detected QTLs in two or more trials, the other QTLs were environment-specific. These results underscore the significant genotype x environment interactions found for most of the traits. Several genomic segments affecting multiple traits were identified and were in agreement with correlation among traits. All QTLs identified for FRY were found associated with either component traits of productivity or architecture traits. This study suggests that QTLs for plant architecture can be used to improve productivity. However an exhaustive search and analysis of QTLs controlling architecture is required before marker-assisted selection (MAS) for increasing productivity can be initiated.