The objective of this study was to formulate a prioritization index to evaluate vegetative containment on a steep road cut alongside a highway connecting the towns of Viçosa and Araponga, Minas Gerais State, Brazil (20º44' latitude S; 42º50' longitude W). Treatments consisted of nine combinations of a grass (Brachiaria brizantha, Brachiaria decumbens or Brachiaria ruziziensis) and a legume (Stizolobium aterrimum, Cajanus cajan or Dolichos lab lab). In developing the index, first step was to choose, among the analyzed characteristics, those most appropriate to select combinations in the revegetation process, thereby establishing frequency by means of a weighted matrix. Upon concluding this exercise, the frequency assigned to the characteristic represents its individual importance in the overall consortium response to treatments. The final score for each consortium is the sum of the product of the frequencies multiplied by the relative values of the characteristics that were experimentally determined. Upon calculation, the indices were statistically analyzed, applying the Scott-Knott cluster test. It may be concluded from the joint standardized index that Stizolobium aterrimum was the most appropriate legume in the revegetation process among the three tested, with a mean value of 28.03 in a scale up to 30. Cajanus cajan showed intermediate response (26.52) and Dolichos lab lab was the least successful (23.08). Similar growth characteristics for the brachiarias allow for alternative responses, depending on the associated legume. Therefore the Brachiaria brizantha was indicated as best for association with Stizolobium aterrimum and Brachiaria decumbens best in consortium with Cajanus cajan.
species selection; grasses; legume; degraded land reclamation
