Evaluation of tree leguminous species seedlings introduced into pastures through multidimensional variance analysis

Abstract

The introduction of unprotected young plants of 16 leguminous tree species into Brachiaria decumbens pastures under grazing was evaluated in three different moments of the year at the Seropedica county, RJ. Nine variables related to length and number of sprouts were used for the evaluations before and after animal grazing. The null hypothesis was rejected by the four Manova tests. Thus, the results of this experiment were analysed by means of multidimensional variance analysis in order to better explore the combined information of the variables. Difference among the 48 treatments means for each variable, calculated by Bonferroni confidence intervals, revealed that Jurema Preta (Mimosa tenuiflora) presented the greatest length and the highest number of sprouts after grazing, the shortest differences of length and number of sprouts as grazing and the greatest number of sprouts before animal grazing at the 3rd and 4th evaluation. The results related to the variables after grazing obtained from Jurema Preta confirm its recommendation for the region as the tree legume with the highest chances of growth whether introduced into pastures under grazing without protection.