Dry matter intake (DMI) of Tanzania grass under grazing conditions was estimated using 24 Holstein-Zebu and Gir cows from the in vitro dry matter digestibility of extrusa samples (esophageal fistulated cows) and fecal output. The fecal output was estimated using chromium mordant and a non-linear model. The pasture was managed in a rotational system with three days paddock occupation and 39 days of resting period. The stocking rate was 2.0 cows/ha during the rainy season. Four different equations based on in situ degradation characteristics were used to predict DMI: DMI = -1.19 + 0.035 (a+ b) + 28.5c (1), DMI = -0.822 + 0.0748 (a+ b) + 40.7c (2), DMI = -8.286 + 0.266a + 0.102b +17.696c (3) and DMI = [%FDN MS]* [FDN intake] / [(1-a-b)/K P +b/(c+ k p)]/24] (4). The DMI predicted from the equations, in general, underestimated the results obtained with the non-linear model (9.6 kg DM/cow/day). The DMI data using the equations (2) and (4) (6.2 and 6.0 kg of DM/cow/day) were similar and different from the results obtained using equations (1) (12.7 kg DM/cow/day) and (3) (8.1 kg DM/cow/day). All the DMI were underestimated in relation to results from the non-linear model, except those from equation (1). The prediction of tropical forages intake, under grazing condition using equations based in degradation characteristics constitute an important tool for these evaluations. Eventhough, more studies need to be done to validate these equations in practice.
degradability; intake; mordanted chromium; pastures; prediction equations; tropical forages
