Silage produced with the entire corn plant (Zea mays L.) has been an increasingly widespread alternative as bulky food for ruminants. One of the main care in crop management with a view to silage production is related to fertilization. In tropical soils (such as those in the south of the state of Pará) phosphorus is the most limiting nutrient for agricultural production. Because of the above, the present work aims to know the agronomic performance of forage corn cultivars in high and low phosphorus in the South of the state of Pará. For this, two competition trials of maize cultivars under contracting conditions of phosphorus (P) were carried out: high P (100 kg ha-1 from P2O5) and low P (50 kg ha-1 from P2O5), using as a source the simple superphosphate (18% from P2O5). The experimental design was randomized blocks with fourteen treatments and three replications. The treatments were composed of fourteen cultivars, eleven of these were the hybrids: AG1051, AG8088, BM3061, BR2022, BR205, BR206, BRS3046, CR-120, M274, ORION, and PR27D28, and the other three were open pollination populations (PPA): AL BANDEIRANTE, ANHEMBI, and CATIVERDE. The variables evaluated were: height of insertion of the ear (AE) and plant height (AP), total weight per plant (PPT), the weight of the cob with straw (PESP), and weight of the ear without straw (PESP). The medium was compared by the test of groups of Scott and Knott, the 5% of significance, using the program SISVAR. Considering the overall performance of the cultivars, in all traits evaluated in the two trials, the sum of their results points to the cultivar BRS3046 as the most promising for silage production. The high test P, as expected, has been used about the low P.

Campus. The design used was randomized blocks, with ten treatments and three replications, where the treatments are represented by ten maize cultivars. The variables were measured when the plant reached R6 stage. For the study of genetic divergence, the following variables were analyzed: The number of grains per row (NGPF), number of rows per spike (NFE), plant height (AP), the height of the spike (AE), moisture (UMID), grain yield (kg). The Tocher method was used to verify similar groups; for the verification of genetic distance, there was the foundation through the generalized distance of Mahalanobis (D²) and Singh's criterion for quantifying the contribution of the variables evaluated. The characteristics of grain yield and number of grains per row were the ones that most contributed to the genetic divergence for treatments with Azospirillum and without Azospirillum. The crosses between cultivars aiming at promising combinations to obtain strains are BM3051 X CATIVERDE with Azospirillum and BM3051 X CATIVERDE without Azospirillum.