RESUMO
Our objectives were to evaluate strategies to improve productivity and economic returns from beef and dual-purpose cattle systems based on data collected on one dual-purpose (Bos taurus x Bos indicus) and two beef (Nellore) cattle farms in the western Amazon region of Brazil. Forage chemical composition and digestion rates of carbohydrate fractions of grazed Brachiaria decumbens and Brachiaria brizantha cv. Marandu grasses and Pueraria phaseoloides (tropical kudzu) legume were measured monthly during a 9-mo period from the end of one dry season to the end of the subsequent rainy season. Measurements of milk and growth responses to grazing these forages were used to predict animal productivity responses to dietary nutrient availability throughout an annual cycle. The ME available for gain in our simulations was always more limiting than metabolizable protein. The predicted ME available for gain was 0.50 kg/d for steers grazing B. brizantha and 0.40 kg/d for finishing steers grazing B. decumbens. Grasses contained more NDF and neutral detergent insoluble protein and less ME (P < 0.05) in the rainiest months than in the less rainy season, which resulted in 20% less predicted weight gain by growing steers (P < 0.05). Supplementation with sorghum grain was required to increase milk production and growth by 25 or 50% per animal, respectively, but this strategy was less profitable than current forage-only diets. Greater productivity of land and labor from higher stocking indicated greater net margins for beef production, but not for milk. This study suggested that more intensive beef production by judicious fertilization of grass-legume pastures and greater stocking density is the preferable strategy for owners of these cattle systems to improve economic returns under current conditions. It also might help decrease the motivation for additional forest clearing.