There were strong reductions in additive effects by two QTL located on chromosomes 1 and 6, and one QTL on chromosome 10. When protein content was conditioned on oil content, one of five QTL with reduced effects

on protein content was detected, and one new QTL was identified on chromosome 2 (Table 4). When starch content was conditioned, all five unconditional QTL for protein content were detected and three new QTL explaining 3.5% to 4.1% of the phenotypic variation for protein content were found. When starch content was conditioned on BGB324 in vivo kernel oil content, none of QTL showed significant effects and four additional QTL accounting for 3.2% to 6.3% of the phenotypic variation were identified (Table 5). When starch content was conditioned on protein content, only four of eight QTL were detected with slightly reduced additive effects. In addition, four new QTL were detected, accounting for 3.4% to 12.4% of the variation in starch content. In summary, more than half of the unconditional QTL for

each measured trait were not detected or showed large reductions, when conditional QTL mapping were performed. These results check details suggest that there is a strong genetic association among oil, protein and starch content in maize kernels. We detected 9, 5 and 7 unconditional QTL for oil, protein and starch content in the presently investigated RIL population, one of whose parents involved BHO background. In the early generations of this RIL population (F2, F3 and F2:3), a total of 26 QTL were detected (15 for oil, 6 for protein, 5 for starch) [15] and [16]. Combining the present and previous Etomidate studies using B73 × By804 segregating populations [15], [16], [17] and [18], 10, 4 and 3 QTL were detected in over at least two generations. In contrast, about 66, 66 and 65 loci for oil, protein and starch content had been reported in six different populations generated from IHO germplasm [7], [8], [9], [10], [11], [12] and [13]. Furthermore, QTL for three quality traits detected in IHO and BHO populations

were compared using the IBM neighbor genetic map (http://www.maizegdb.org/) as a bridge. For oil content, about 20 QTL were detected in both germplasms. However, the strongest QTL in IHO germplasm was detected in Bin 6.04, and QTL in Bin 1.04 had the largest effect on oil content in BHO germplasm. For protein and starch content, most of the QTL in BHO germplasm coincided with IHO germplasm except QTL proc9-1, which explained 7.7% of the phenotypic variation for protein content on chromosome 9 (Bin 9.04–9.05). These results suggest that there might be many different loci for maize kernel composition in different maize germplasms in spite of the positional consistency of QTL for oil, protein and starch content across different maize populations. Oil, protein and starch are major chemical components of maize kernels.