Background

Some of the rice germplasm users requested seeds of specific accessions with known traits of interest for breeding and research, while others request trait donors without specifying the accessions. Due to financial and technical constraints in characterizing the genetic variation of a large collection of rice germplasm, rice breeders in Africa exploited the genetic potential of only less than 10% of our collection conserved at our genebank as parents in their breeding program. Most users request seeds of fewer accessions that are known to them, which may be due to (1) lack of publicly available trait-specific gene pools (trait donors) to the user; (2) inadequate linkages between our genebank and germplasm users in the continent, and (3) limited capacity of the NARS breeding programs in exploiting the potential of our collections. In some cases, breeders extensively used only a few well-characterized/evaluated accessions as parents of various improved varieties²⁷, which not only tend to bring redundant genetic value to a breeding program but also reduces the overall genetic gains and increases the risk of devastation in cases of major diseases and pests outbreaks. To avoid excessive use of a limited number of parental line combinations in future rice breeding programs and increase genetic gains, therefore, there is a strong need in investing more resources on molecular characterization using high-density markers, creating subsets that capture the genetic diversity of every species (subsetting), evaluation of the subsets, and using the evaluation data for trait donor identification and gene discovery, which are described in the next sections.