EVALUATION OF FECAL INDICATORS TO PREDICT DIGESTIBILITY AND VOLUNTARYINTAKE OF DICHANTHIUM SP BY CATTLE

In vivo digestibility trials involving creole bull calves were carried out in order to evaluate the potential of faecal indicators for asses sing organic matter digestibility (dMO) and organic matter intake (MOI ). The evaluated faecal indicators were crude protein (MATf), neutral detergent fiber (NDR), acid detergent fiber (ADFf), residual nitrogen in ADF (NADFf) and in vitro dry matter digestibility (DIV). A continuo us digestibility trial lasting 53 days was conducted with Dichanthium sp harvested from 20 to 73 days of regrowth. Six creole bull calves (2 56 +/- 32 kg) housed in metabolism cages were led ad libitum twice dai ly. Refusals and faeces were collected and weighed every day for each animal. Dry matter (MS), MAT, NDF, ADF, NADF content and DIV were dete rmined by standard procedures from forage and faeces dried samples. Me an values were calculated per week (5 days) period for each animal. MO I was 4,88 kg/day and dMO ranged from 0.68 to 0.55. In our conditions, the best predictor of dMO was MA Tf content. Linear quadratic and hyp erbolic functions were of similar accuracy (rsd = 0.025). NDRf ADPf an d NADFf were less accurate. Faecal MAT output (g/day) was the best pre dictor of Mal (rsd = 240 g). The best accurate regressions predicting dMO or MOI were used to estimate digestibility and intake of tethered creole heifers grazing Dichanthium sp pastures during three trials whe re daily body live-weight gain were different (740, 430 and 140 g). On ly the hyperbolic model predicting dMO (dMO = 0.983 - 4.002/MATf, ETR = 0.023) allowed estimates of MOI and energy intake consistent with th e average daily gain and energy requirements of the grazing heifers in all trials. This model should be preferred.