2021 Volume 23 Issue 1 Pages 16-27
Many papers about rice genes related to yield are published every year, but few of these genes have been used in practical rice breeding. Therefore, we generated or received near-isogenic lines (NILs) of nine alleles related to yield in the genetic backgrounds of Japanese rice cultivars, and performed simultaneous comparative cultivation tests. Significant increases in brown rice yield were detected in the NILs for qCTd11Takanari, SD1DGWG, and a combination of TGW6Kasalath and SD1DGWG in the genetic background of ‘Koshihikari’ and a NIL for DEP1Ballila in the genetic background of ‘Akidawara’. These alleles and QTLs may have the genetic potential to improve rice yield, even in the genetic background of Japanese cultivars. On the other hand, an increase in the brown rice yield of NILs for qLIA3Takanari, GPSTakanari, GS3Oochikara, GW2BG1, and GN1ATakanari was not detected in the ‘Koshihikari’ genetic background probably due to small direct effects to increase sink size by these alleles and QTLs or a relative reduction of their source abilities to the total sink sizes increased in these NILs. In addition, analysis using multiple NILs in the ‘Koshihikari’ genetic background showed a negative correlation between panicle weight and panicle number. This indicated that an increase in the number of panicles did not subsequently increase yield because of the tradeoff to reduce the weight of one panicle and the thousand grain weight; this was considered the main reason why GW2BG1, GS3Oochikara, and GN1ATakanari, which were expected to increase sink capacity, did not increase yield. We considered that the alleles increasing sink capacity are not sufficient to increase yield in the genetic background of Japanese cultivars and need to be combined with alleles, such as qCTd11Takanari, expected to improve the source ability by increasing photosynthesis per unit leaf area or alleles to improve canopy structure including SD1DGWG and DEP1Ballila.