Estimation of the indigestible fiber in different forage types.

The role of indigestible NDF is essential in relation to OM digestibility prediction, total tract digestibility, rumen fill, passage rate, and digestion kinetics. Moreover, the truly indigestible NDF (iNDF) represents a core point in dynamic models used for diet formulations. However, despite its wide possible applications, few trials have been conducted to quantify iNDF and even fewer to investigate whether or not it is consistent among different forage sources. The objective of this study was to predict the iNDF by measuring the residual NDF after 240-h in vitro fermentation to determine the unavailable NDF (uNDF) within and among various forage types. Finally, a mathematical approach was investigated for the estimation of the uNDF fraction. In all, 688 forages were analyzed in this study. This pool included 122 alfalfa hays, 282 corn silages, and 284 grass hays. Values of uNDF varied among different forages and within the same type (22.7% ± 4.48%, 20.1% ± 4.23%, and 11.8% ± 3.5% DM for grass hay, alfalfa hay, and corn silages, respectively). The relationship among uNDF and ADL was not constant and, for grass hay and corn silage, was different ( 0.05) from the 2.4 × lignin value applied by the traditional Chandler equation. The observed uNDF:ADL ratio was 3.22 for grass hay and 3.11 for corn silage. Relationships among chemical and biological parameters and uNDF were investigated via simple and multiple regression equations. The greatest correlation with a single variable was obtained by ADL and ADF when applied to the whole data set ( = 0.63). Greater coefficients of determination resulted from a multiple regression equation for the whole data set ( = 0.80) and within each forage type ( = 0.65, 0.77, and 0.54 for grass hay, alfalfa hay, and corn silage, respectively). In conclusion, a regression approach requires specific equations and different regression coefficients for each forage type. The direct measurement of uNDF represented the best approach to obtain an accurate prediction of the iNDF and to optimize its specific purpose in dynamic nutrition models.