Yields of B. decumbens cultivar Basilisk and B. brizantha cultivars Marandu and Xaraés were evaluated under different shade levels. A completely randomized block split plot design was used, with the shade levels as plots (0, 50, 70%), and the species and cultivars as split-plots. Seeds were germinated on trays and three seedlings were transplanted to each pot. Four harvests were made on each split-plot. Prior to each harvest, plant height, number of live tiller per pot and the leaf area of four leaves per pot were recorded. After harvesting, the plants were separated into leaves, stem + sheaths and dead matter to determine dry matter yield. Total dry matter yield was calculated for each harvest and for the sum of harvests. In the last harvest, root dry matter yield was also evaluated. Significant interactions were unfolded adequately. The means of the qualitative factor (forages) were compared by the Tukey test and the quantitative factor (shade level) was submitted to linear regression analysis. With the exception of the mean yield in four harvests and stem and dead matter yield, significant interaction was detected for all other variables. Regardless of the shade level, the Xaraés cultivar presented the highest values in most of the variables analyzed, while B. decumbens presented the lowest yield in all levels of shade. Regarding leaf percentage, cv. Xaraés revealed the highest results, followed by cv. Marandu. Regardless of the forage grass, the shade level affected directly the number of tillers/plant, root dry matter yield, leaf area and SPAD value.
brachiaria grass; marandu grass; silvipastoral; xaraes grass
