Completing an Expert System That Will Provide Site-Specific Nutritive Values for Feeds

CRIS # 3655-31000-018-00D

Timeline: September 17, 2002 to September 26, 2007

Personnel:
Scientists: David Mertens (0.7), Richard Muck (0.6), Glen Broderick (0.3), John Grabber (0.2)
Total SY = 1.8
Technical support: Diane Amundson (1.0), Ursula Hymes-Fecht (1.0)
Temporary support: Grad Students (1.0), Undergrads (1.5), Post-docs (1.0)

Importance:
Currently, routine nutritive evaluation of feeds is based primarily on chemical analyses, but the relationship between chemical composition and nutritive value is variable. Other factors, such as crop growing conditions; methods of harvesting, storing and processing the feed; the ration in which it is fed; and the animal receiving it affect the specific nutritive value of a feed. Thus, chemical analysis alone results in inaccurate feeding programs that affect sustainability and profitability, and also contaminate the environment when excess nutrients are fed.

Objectives:
• Comprehensive reference database of feed information for the expert system.
• Relationships between crop growing conditions and nutritional value of forages.
• Improving relationships that adjust nutritive value during harvesting, storage and preservation.
• Fundamental relationships between lignification or ruminal pH that alter digestible NDF and energy value.
• Evaluating the feed information expert system

Projects:
• Expert system software that integrates information to estimate site-specific nutritive value.
• Impacts of environmental and weather conditions on the nutritive value of corn silage.
• Relationship of ensiling techniques on the preservation and nutritive value of silages.
• Factors affecting the kinetics of digestion and fermentative gas production of feeds.
• In vivo and in vitro digestibility of corn silage as influenced by starch particle size.
• Intake and nutritive value of alfalfa cut in the AM or PM.
• Methods for measuring physically effective NDF.
• Rapid methods of analysis for extent of maceration, digestion kinetics, intake, and digestibility.
• Models for predicting ruminal pH and turnover of liquids and solids.

Impact:
A 5% underestimate of the energy value of a diet raises feeding costs by $0.03 per cow per day. An overestimate of nutritive value is more costly ($0.09-0.12 per cow per day) because of lost animal performance and well-being. Eliminating these inaccuracies for 25% of the dairy cows in the US would improve dairy industry profitability by about $40 million annually. In addition to improving animal diets, improved accuracy of ration formulation would significantly reduce excess nutrient excretion and adverse environmental effects from manure disposal.