Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: Multiple Regression Equation for Predicting Relative Feed Value of Grass Hays

Authors
item Moore, J - UNIVERSITY OF FLORIDA
item Burns, Joseph
item Fisher, Dwight

Submitted to: American Forage and Grassland Council Conference Proceedings
Publication Type: Proceedings
Publication Acceptance Date: June 13, 1996
Publication Date: N/A

Interpretive Summary: In 1972, the American Forage and Grassland Council (AFGC) developed an overall index of forage quality, called Relative Feed Value (RFV), which combines both intake and digestibility. In addition, AFGC proposed using Acid Detergent Fiber (ADF) and Neutral Detergent Fiber (NDF) analyses of forages to predict digestibility and intake, respectively. The National lForage Testing Association (NFTA) certifies commercial laboratories for ADF and NDF analysis, and uses equations to convert ADF and NDF to estimates of RFV. The RFV is used to establish the price of hays at hay markets throughout the Midwest. Research has shown, however, that NDF is not related to intake in many cases, thus suggesting that the conversion equation for predicting intake should be reevaluated, especially for grasses. Both temperate and tropical grasses were fed to cattle for intake measurement and to sheep for digestibility measurement. Samples were analyzed using a variety of methods. The NFTA equations gave digestibility values which were similar to observed values, but intake was underestimated. As a consequence of the error in intake estimates, estimates of RFV were generally lower than the observed values, particularly for high-quality grasses. In addition, there was a lot of variation between observed values and the estimates of both intake and digestibility. New conversion equations were developed using two or three analyses, including in vitro digestion. The new equations were more acceptable than the NFTA equations for grasses, and should be considered for use in establishing RFV in grass testing programs.

Technical Abstract: Regional Project S-45 data were used to develop multiple regression equations for predicting in vivo dry matter (DM) digestibility (DMD), voluntary DM intake (DMI), and Relative Feed Value (RFV). Grass hays were grown and fed in eight states and included 25 temperate and 27 tropical hays. Hays were fed to sheep for measuring DMD, and to cattle for DMI. Samples were analyzed in 11 laboratories for crude protein (CP), acid detergent fiber (ADF), neutral detergent fiber (NDF), and in vitro digestion (IVD) by several procedures. Data from individual laboratories were used to compute prediction equations. No single laboratory analysis gave acceptable predictions of DMD or DMI, but predictions from multiple equations were acceptable. For predicting grass hay RFV, multiple regression equations are more acceptable than are equations presently used by the National Forage Testing Association (NFTA).

Last Modified: 7/23/2014
Footer Content Back to Top of Page