Significant genotype x environment interaction for quantitative traits, such is grain yield, reduces the usefulness of genotype means, over all environments, for selecting superior genotypes. AMMI model is a valuable statistical tool in identifying systemic variation contained in the interaction effect. Obtained data could be applied in maximizing yield potential in every environment based on both narrow and wide genotype adaptability, without the necessity of developing breeding programs for smaller targeted environments. Precise assortment of superior genotypes, with the assistance of AMMI model, leads to the better recommendation of newly bred hybrids, and thus increasing maize grain yield in a targeted environment. In this research genotype x environment interaction and yield stability of 36 maize hybrids of FAO 300-700 maturity group was investigating. The trial was set according to Randomized Complete Block Design (RCBD). Data were processed in order to obtain average estimates of grain yield, and yield stability was assessed by the method of AMMI analysis. The highest average grain yield was achieved in 2011 (11.62 t/ha), and the lowest in the most stressful and dry 2012 (6.90 t/ha). In the region Loznica L2 the highest average yield was noticed (13.81 t/ha), while at L7 (Sremska Mitrovica) average grain yield was the lowest (6.97 t/ha). Results of AMMI analysis gave precise recommendation for production of maize hybrids in certain environments, by determining winning areas of hybrids H20, H11 and H36. Medium early maturing and high yielding hybrids (H11 and H20) are therefore considered more favorable for production in environments with lower precipitation, while high yielding and more stable hybrids H21 and H35 are suitable for a wider range of environments. Hybrid H36 (FAO 700) showed its full potential at L2, and L3 which did not suffer from a lack of moisture. This hybrid also expressed its best potential in environments with favorable conditions.

<![CDATA[In maize (Zea mays L.) breeding programs, it is very important to define a genetic distance of inbred lines and to establish criteria and biometric methods for the satisfactory germplasm classification. A total of 29 inbred lines from Maize Research Institute "Zemun Polje" (MRIZP) breeding program, Belgrade, Serbia, were used to compare similarities obtained by morphological (according to UPOV - Union Internationale pour la Protection des Obtentions Vegetales) and molecular (simple sequence repeats, SSR) markers. The aim was to assess discriminative power of applied markers in the separation of related lines, and to determine the possibility of their application in choosing parental lines for breeding programs. Application of different marker types resulted in adequate separation of inbreds into divergent groups, which facilitates the choice of parental lines. However, there were still many inconsistencies between the field and laboratory results. The quality of information from morphological markers was improved by the application of the appropriate descriptor, measurement scale and the biometric method. According to the correspondence analysis, increased number of SSR markers will enable more reliable results. Regression analysis of morphological visual assessment (MVG) and SSR similarity of individual inbreds pointed out that in some cases morphological, and in others molecular markers more accurately reflected known pedigree information of tested maize lines. In the early generations of maize inbred lines testing, we propose utilization of information resulting from morphological markers, according to UPOV descriptor. However, application of adequate number of molecular (SSR) markers has an economic justification on a smaller number of elite lines in the later generations of line testing.]]>

In the second half of 20^th century the awareness of importance of landraces for the future, led to organized collecting missions for numerous plant species. A total of 2217 maize (Zea mays L.) landraces, collected in the former Yugoslavia, are stored at Maize Research Institute (MRIZP) gene bank. During 2014, new collecting missions were organized in the eastern and western parts of Macedonia. According to collecting site and kernel type, 14 samples from the Faculty of Agricultural Sciences and Food, R. Macedonia were chosen for the comparison and identification of possible duplicates, through coupling with the 16 MRIZP gene bank accessions from the same area and kernel characteristics. Phenotypic characterization was done for 21 traits according to International Board for Plant Genetic Resources descriptors for maize. The Principal Component Analysis (PCA) identifies five PCs with Eigenvalue > 1, explaining 80% of the total phenotypic variation. The most discriminative traits with the strongest positive associations were tasseling and silking dates, plant height, leaf length and ear length. Compared to the ex-situ populations, the number of alleles and the number of specific alleles, showed a significant decrease in the in situ populations. Twelve unique alleles were detected in samples from MRIZP gene bank, and only four were found in new Macedonian samples. Cluster analysis of morphological and molecular markers distinguished groups of maize accessions with distinctive morphological traits and genetic profiles that will be useful for conservation, and management of gene bank collection, as well as for possible utilization in breeding.

In the process of plant breeding the application of relevant breeding criteria is very important. The Path analysis is broadly applied with the aim to define yield components that mostly determine the yield and that can be used as quality breeding criteria. However, the significance of revealed relationships between yield and yield components can be affected by various factors, such as diverse genetic material that is observed, traits included into analysis, environments in which the material is observed, as well as, the applied statistic approach to determine the nature of the relationships itself. The interrelationships of yield and yield components of 15 commercial maize hybrids were observed using the Path and factor analyses. According to results of Path analysis, plant height, ear diameter and grain moisture had highly significant genetic and phenotypic direct effects on grain yield. At the same time, factor analysis points out significant effects of two factors on grain yield. Factor 1 was mostly determined by ear length and number of kernels per row, while grain moisture content, ear and cob diameter mostly determined Factor 2.