Equations for predicting elephantgrass (Pennisetum purpureum, Schumack) dry matter intake (DMI) by Holstein x Zebu lactating dairy cows under grazing were developed by stepwise regression analysis. The dataset consisted of results of trials carried out at Embrapa Gado de Leite (Coronel Pacheco, MG, Brazil). The independent variables related to animal factors (days in milk; milk protein, fat and total solids contents and yields; live weight; live weight at calving; milk and fat-corrected milk yield; Holstein x Zebu gene fraction; order of parturition); management factors (occupation day of paddock; forage allowance and availability; resting periods); environmental factors (season of the year; rainfall) and feed factors (in vitro digestibility and parameters of chemical composition of elephantgrass herbage and sugarcane - Saccharum officinarum (L.) plus 1% urea; supplement intake - concentrate and sugarcane plus urea; fecal crude protein, neutral and acid detergent fiber concentration). Logarithmic and quadratic functions for some variables were also considered. The main independent variables selected at the best-fit regressions models (R2 = 65.2-67.0%) for elephantgrass herbage DMI were roughage supplement (sugarcane plus urea) DMI; elephantgrass in vitro digestibility; rainfall; fat-corrected milk yield; live weight (or alternatively live weight at calving) and concentrate intake that showed substitution effect to intake of elephantgrass in supplementing grazing cows.
dry matter intake; independent variables; intake models; lactating cows; prediction equations; supplementation
