| Popis: |
Assessment of forage nutritive value facilitates its appropriate inclusion in diets for ruminants. The objective of the present study was to determine the precision with which laboratory and in situ measurements could predict the digestible organic matter in dry matter (DOMD) in vivo of mountain forages, using multiple linear, step-wise regression analyses. A joint research was carried out between the Federal Research Institute for Agriculture in Alpine Regions (BAL Gumpenstein, Austria) and the Department of Animal Science of Padova University (Italy). The data set included 66 native mountain forages for which in vivo digestibility data were available. Forage harvesting methods (hay, silage (SIL) and fresh grass) were forced to enter as dummy variables in the equations. Four sets of independent variables: Weende constituents (crude protein, ash and crude fibre), Van Soest fibre constituents (NDF, ADF and ADL), in situ 48 h DM degradation and in vitro techniques (NDF-cellulase, pepsin-cellulase (PEPCEL), Tilley-Terry and gas production HFT) were submitted to a step-wise regression analysis. In the first stage of the regression, the four sets of variables were separately analysed and the best equation was obtained with the in vitro pepsin-cellulase procedure: DOMD = 44.90 − 5.92 HAY −7.61 SIL + 0.35 PEPCEL (n = 66; mean = 57.68 ± 5.06; RSD = 3.43; R2 = 0.540). The second stage of regression implied the subsequent inclusion of the sets following a criterion based on the higher degree of complexity of the analytical methods (Weende plus Van Soest plus in situ plus in vitro). The resulting final equation was: DOMD = 19.85 − 5.94 HAY − 4.97 SIL − 0.70 ASH + 0.25 NDF + 0.64 PEPCEL (RSD = 2.86; R2 = 0.681). The equation, which increased the precision in predicting DOMD in vivo, requires only a restricted number of laboratory determinations and may be suitable for a widespread utilisation by mountain farmers. |
