Analysis of rice panicle traits and detection of QTLs using an image analyzing method

摘要

Controlling the grain number in rice panicles is essential for increasing the yield potential. Because this trait depends on some complicated panicle branching structures, we prepared a newly developed software, PASTAR (PAnicle STructure Analyzer for Rice) and PASTA Viewer, to automatically extract values of the length and number of various branches and grain number from scanned photographic images of panicles. Furthermore, PASTA Viewer could draw panicle branching patterns based on actually measured values. Using this image analyzing method, we analyzed some extra numerous grain rice cultivars and performed quantitative trait locus (QTL) analysis for 18 panicle traits in the F-2 segregated population derived from a cross between a japonica rice cultivar, Koshihikari, and an extra numerous grain cultivar, NP-6, classified as a "New Plant Type". The results revealed that QTLs enhanced by NP-6 alleles for some panicle traits were concentrated on a few regions of chromosomes 1, 6 and 8, although some single QTLs were also detected in various chromosomal regions. This suggests that the NP-6 panicle and extra numerous grain traits are predominantly controlled by a few gene loci acting pleiotropically, along with some partially effective panicle structure genes.