Rice Plant Breeding

Objective

The objective of our research is to implement a marker assisted selection program which uses a physiologically-based crop simulation model to define which combination of phenotypic traits, and therefore genes, to select when breeding for higher grain yield. The advantage of the proposed approach is that it would provide greater focus to rice breeding programs over what is achievable using either a conventional plant breeding program or a conventional marker assisted breeding program, both of which are limited in their ability to identify the "best" genetic or phenotypic combination.

Currently, the rice breeding program at Beaumont, Texas evaluates about 30,000 selections each year, with about 25 to 30 new pure lines evaluated for their potential as varietal releases each year in regional trials. At different stages of the selection process, measurements are taken of phenotypic traits for maturity, yield, lodging, grain quality, and disease resistance. The culmination of this process is the release of a new cultivar about every 5 to 7 years, at a cost of about $7-10 million per cultivar.

Numerous lines possessing desirable genes are discarded during the repeated selections because they are not recognized as being desirable within their existing phenotypic background.   Similarly, numerous lines possessing undesirable genes are maintained long after considerable investment has been expended on the selection process.  A pressing need is technology that would enable a major increase in the speed with which desirable lines can be identified. The proposed merging of marker assisted selection with phenotype modeling would allow rapid definition and selection of rice ideotype(s) with the greatest economic benefit to the Texas rice industry.

Document Author:

Photos by:

Revised: