Abstract: 【Objective】 Kernel row number (KRN) is directly related to maize yield. QTL for KRN was mapped to dissect the genetic basis of KRN, which was helpful to the improvement of yield performance in maize.【Method】In order to dissect the genetic control of KRN of Four-rowed waxy corn, a landrace collected in southern Yunnan, the segregation populations were developed by reciprocally crossing the inbred line derived from Four-rowed waxy corn with Nong531, an inbred with 18-22 rows. The reciprocally crossed F2:3 populations derived from F2 by single seed descend method were separately phenotyped in Beijing (BJ) and Henan (HN). Trials were conducted in randomized complete blocks with two replications. Totally 173 polymorphic SSR primer pairs obtained from the maize genome database (http://www.maizegdb.org) were used to genotype the lines and to construct linkage maps. QTL and their interactions for KRN were detected with ICIM (QTL IciMapping Version3.2 Software) and CIM (QTL Network2.0 Software). The phenotypic variation explained by the major QTL was estimated with SAS GLM software. 【Result】 Phenotypic analysis showed that the average KRN of Four-rowed waxy corn and Nong531 was 4.0 and 19.2, respectively, and the KRN of the reciprocally crossed F2:3 families ranged from 4.0 to 17.4. In two environments, a total of 12 QTL were detected with ICIM, which were located on eight chromosomes. All alleles having negative effect on KRN were from Four-rowed waxy corn. A total of five major QTL could be detected in both environments. qKRN2-1, which was the QTL with the largest effect, explained 18.48% of the phenotypic variance. The other four QTL, which were qKRN4-2, qKRN5-1, qKRN8-2 and qKRN9-1, explained 11.58%, 13.55%, 16.91% and 9.66% of the phenotypic variance, respectively. Five QTL were detected with CIM, four of which were in accordance with the major QTL detected with ICIM, that were qKRN2-1, qKRN4-2, qKRN8-2, and qKRN9-1. Cumulatively, the five major QTL explained 51.5% (BJ) and 54.0% (HN) of the phenotypic variance in the two environments with the GLM model, respectively. Correlation analysis showed that, in the two environments, correlations between the estimated values of phenotypic variance and observed values in F2:3 families were highly significant. In addition, two pairs of QTL located on three chromosomes (Chr. 2, Chr. 4 and Chr. 9) represented epistasis, explaining only a small part of the phenotypic variance, which were 2.90% and 1.80%, respectively. 【Conclusion】 Four-rowed waxy corn possessed a number of alleles with KRN-reducing effect and it could be a good resource in genetic analysis of maize KRN. The five QTL located in bin2.04, bin4.09, bin5.04, bin8.05 and bin9.03 could be considered as the major candidates for fine mapping and cloning in the future.

Key words: Zea mays L. , Four-rowed waxy corn , kernel row number (KRN) , quantitative trait loci (QTL) , genetic dissection