Submitted to: Crop Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: December 16, 2000
Publication Date: N/A
Interpretive Summary: Grain shape is a fundamental quality characteristic of rice and is used to define U.S. market classes. Grain shape and size are also correlated with yield (quantity) and percent whole grains after milling (quality), thus affecting farmers' incomes. We identified 21 chromosomal regions in rice that contain genes affecting rice kernel shape. These genes were found through genetic association with molecular markers. More genes were identified with analysis of simple, single-dimension traits (kernel length, width, and thickness) than with compound calculated traits such as shape (defined in rice as length to width ratio), and volume which we estimated from length x width x thickness). It was also apparent that the complex trait data did not estimate the gene locations and effects as accurately as analysis of length, width, and thickness per se. Particularly in cases where a length gene was linked to an opposing gene for width, the analysis of shape or volume indicated the presence of a single gene located between the genes for length or width; as if the length and width effects were averaged together. Grain weight genes that were previously mapped in this population resided with the shape and volume genes, between the more precisely identified genes for length and width. This is an example of how analysis of component traits can greatly enhance efforts to identify genes affecting complex traits such as grain weight and volume. In further studies, aimed at finding genes for even more complex traits such as yield and grain quality, scientists would be wise to first identify simple, underlying component traits.
Grain dimensions are fundamental quality characteristics of rice and are used to define market classes. Grain dimensions are also correlated with yield (quantity) and percent whole grains after milling (quality), thus affecting farmers' incomes. To examine the genes affecting kernel dimension, we conducted RFLP analysis with 175 RFLP markers and kernel measurements from 308 recombinant inbred lines derived from 'Lemont' x 'Teqing'. Lemont is a Southern U.S. long-grain cultivar with good milling quality. Teqing is a very high yielding cultivar from China with medium- grain shape and poor milling quality. Kernel length, width and thickness were determined from images captured by a digital camera from seed harvested from 1997 and 1998 field plots, two replications per year. Rice industry standards define grain shape as length to width ratio; volume was estimated from length x width x thickness. Transgressive segregation was noted for all kernel dimensions. Twenty-one genomic regions contained QTLs affecting grain dimensions. The QTLs accounted overall for 22% of the total phenotypic variation observed in 1998 for thickness, and from 45 to 67% of the 1998 phenotypic variation observed for length, width, length/width ratio, and volume. Greater within-sample variation was detected in 1997, and QTLs explained less of this increased variation. QTL mapping based on compound traits(shape and volume) did not detect as many QTLs as analyses of length, width, and thickness. In cases where a length QTL was linked to a width or thickness QTL, analysis of compound trait data mislocated a single QTL between the two single-dimension QTLs. This demonstrates how analysis of component traits can enhance identification of QTLs affecting complex traits.