Two experiments, with natural and controlled pollination, were carried out at two sites in the State of Minas Gerais (in Experimental area of Biology Department of Universidade Federal de Lavras - UFLA and UFLA farm, known as "Vitorinha") during 2004/2005, in order to verify and quantify the xenia effects in maize. In both experiments microsatellite markers were used to discriminate kernels derived from crosses and from self-pollination, to detect the xenia effect. Three single cross hybrids were used P 30F90, A 2555 and DKB 333 B, and all possible crosses between them were established, including reciprocal ones. In both trials, composite kernels samples taken from the mid-region of 10 randomly collected ears were evaluated. Based on the evaluated obtained data, through analyses of variance and test "t" for means from unpaired data, the xenia effect on the individual grain weight and the weight of 100 kernels was estimated. The use of microsatellite markers could effectively differentiate the crosses from the selfings in both experiments in all possible combinations of the commercial hybrids used. According to the hybrids involved in the crosses, and the condition pollinator/receptor of each one, xenia effects of different magnitudes were observed. Crosses with 100% allopollen resulted in a 7.3% (2.8 g) increase of the weight of 100 grains compared to those with only 50% allopolen, demonstrating an influence of the greater pressure of foreign pollen on the trait increase. In open-pollinated plants, a mean effect of 12.6% (varying from 7.4 to 16.5%) was observed in the increase of individual grain weight. With 100% foreign pollen, the mean xenia effect was 13.1% on the weight of 100 grains and 8.7% on the individual grain weight, while increments of 15.4% and 16.6% in these traits, respectively, were observed in the most favorable crosses.
maize; xenia effect; single-cross hybrids; grain yield
