Phytic acid (PA) is the main storage form of phosphorus (P) in seeds. It can form insoluble complexes with microelements, thereby reducing their bioavailability for animals. Identification of quantitative trait loci (QTLs) associated with grain PA concentration (PAC) is essential to improve this trait without affecting other aspects of grain nutrition such as protein content. Using a recombinant inbred line (RIL) population, we mapped QTL for grain PAC, as well as grain nitrogen concentration (NC) and P concentration (PC) in maize under two N conditions in 2 yr. We detected six QTLs for PAC. The QTL for PAC on chromosome 4 (phi072-umc1276) was identified under both low-N and high-N treatments, and explained 13.2 and 15.4% of the phenotypic variance, respectively. We identified three QTLs for grain NC, none of which were in the same region as the QTLs for PAC. We identified two QTLs for PC in the low-N treatment, one of which (umc1710-umc2197) was in the same interval as the QTL for PAC under high-N conditions. These results suggested that grain PAC can be improved without affecting grain NC and inorganic PC.

Stalk related traits, comprising plant height (PH), ear height (EH), internode number (IN), average internode length (AIL), stalk diameter (SD), and ear height coefficient (EHC), are significantly correlated with yield, density tolerance, and lodging resistance in maize. To investigate the genetic basis for stalk related traits, a doubled haploid (DH) population derived from a cross between NX531 and NX110 were evauluated under two densities over 2 yr. The additive quantitative trait loci (QTLs), epistatic QTLs were detected using inclusive composite interval mapping and QTL-by-environment interaction were detected using mixed linear model. Differences between the two densities were significant for the six traits in the DH population. A linkage map that covered 1 721.19 cM with an average interval of 10.50 cM was constructed with 164 simple sequence repeat (SSR). Two, two, seven, six, two, and eight additive QTLs for PH, IN, AIL, EH, SD, and EHC, respectively. The extend of their contribution to penotypic variation ranged from 10.10 to 31.93%. Seven QTLs were indentified simultaneously under both densities. One pair, two pairs and one pair of epistatic effects were detected for AIL, SD and EHC, respectively. No epistatic effects were detected for PH, EH, and IN. Nineteen QTLs with environment interactions were detected and their contribution to phenotypic variation ranged from 0.43 to 1.89%. Some QTLs were stably detected under different environments or genetic backgrounds comparing with previous studies. These QTLs could be useful for genetic improvement of stalk related traits in maize breeding.

Association mapping has emerged as a new tool to elucidate complex quantitative trait loci in maize, but there are fewreports about systematic association analysis for the specific SSR markers with agronomic traits of interest in China.We investigated the morphological and genetic diversity and population structure for 76 maize recombinant inbredlines, and then association analysis were further performed between 48 simple sequence repeat loci and 17 morphologicaltraits, consisting of nine ear-related traits and eight other traits. The 48 SSR markers were screened out and furtherclassified into two groups including a group of loci in regions harboring reported quantitative trait loci that affect earshape and a group of markers distributing on the whole genome randomly. The result indicated that the population ofrecombinant inbred lines was structured, showing five subpopulations. Our association results revealed that therewere 82, 59, and 40 significant associations detected by K-test, logistic regression, and both analysis, respectively.When the 17 traits were considered separately, the significant associations between Q-SSRs and E-traits were raised to27.8%, whereas the other groups of combinations ranged between 2.3 and 6.3%. As the proportion of significantassociations is higher among the Q-SSR subset of markers and the subset of traits related to ear shape than those forall of the other combinations, we conclude that this approach is valid for establishing true positive marker-traitrelationships. Our results also demonstrated that association mapping could complement and enhance previous QTLinformation for marker-assisted selection.