Detection and Positional Cloning of a Rice Seed-Length QTL Between ‘Koshihikari' and ‘Nipponbare' (temperate japonica) Using SmartGrain, New High-Throughput Measurement Software

In rice production, seed shape and size are among the most important agronomic traits, because they affect yield and market price. We analyzed quantitative trait loci (QTLs) for rice seed shape using 127 backcross inbred lines (BILs) derived from a cross between two temperate japonica cultivars: Nipponbare and Koshihikari. To measure the small differences among these cultivars and BILs precisely, we used newly developed software called SmartGrain, which can measure a large number of seeds rapidly. The software assessed six parameters almost automatically from scanned images of rice seeds: grain length, grain width, length-to-width ratio, area size, perimeter length, and circularity. QTL analysis was performed using genotype data for 624 markers, most of which were single-nucleotide polymorphism (SNP) markers distributed on 12 rice chromosomes. The QTL analysis detected 13 loci. Among them, four QTLs for seed length, length-to-width ratio, circularity, and perimeter length were detected near a common SNP marker on the long arm of chromosome 11. We confirmed the allele effect of the QTL using Nipponbare–Koshihikari chromosome segment substitution lines (CSSLs). The Nipponbare allele at this QTL increased grain length, resulting in increased grain weight. The QTL was mapped within a ~16-kb region using advanced backcross progenies and a candidate gene was identified.