Current NRC recommendations for dairy cattle provide limited guidance to nutritionists for meeting the fiber and carbohydrate needs of lactating cows. The NRC provide only minimum recommendations for fiber and no accommodation for factors such as physical effectiveness of fiber, interactions with nonfibrous carbohydrates, or animal attributes, which can affect the optimality of dairy rations. To be an improvement, any new system for meeting the fiber requirements of dairy cows must be based on 1) feed characteristics that can be defined and preferably be determined quantitatively using routine laboratory methods and 2) animal requirements that correspond to critical feed characteristics and vary with feeding situation, ration composition, and attributes of the animal. Published data were used to develop coefficients for defining the physical effectiveness or roughage value of feeds and the fiber requirements of dairy cows. Information in this paper is intended to provide practical guidelines for improving current fiber recommendations and to serve as an idealized framework for future research on meeting the fiber requirements of dairy cows. The system is based on NDF as the measure of total chemical fiber in feeds. Adjustments for the effectiveness of NDF in maintaining milk fat production and optimizing ruminal fermentation are based on the particle size and inherent characteristics of NDF that affect chewing activity, ruminal pH, and milk fat production.

Intake and digestibility of feeds by ruminants are influenced by characteristics of the feed, animal and feeding situation. Integration of these characteristics in mathematical models is critical to future progress in forage evaluation and optimal formulation of diets for ruminants. The physiological and physical theories of intake regulation can be described by simple mathematical equations. These equations indicate that intake is a linear function of animal characteristics, such as body weight and production level, and a reciprocal function of feed characteristics, such as fill effect and energy content. Theoretical equations were developed to predict intake when the neutral detergent fiber and energy content of the diet and the energy requirements of the animal are known. The theoretical model also can be used to predict the maximum intake that will maintain a given level of animal production by solving the physiological and physical intake equations at their intersection. Psychogenic intake regulation, which is related to the animal's behavioral response to factors not related to physiological or physical characteristics, can be described mathematically as a multiplier. Digestibility can be predicted by summing the contents of ideal nutritive entities in feeds, which have true digestibilities near 100%, subtracting their associated endogenous losses and adding the variable digestible fiber content. Steady-state models indicate fractional rates of digestion and passage can be used to define ideal nutritive entities and predict digestibility over a range of kinetic characteristics. The steady-state solutions are particularly useful in understanding and predicting the depression in digestibility associated with changes in rates of passage at high levels of feed intake.

Purified corn and wheat starch were added to alfalfa, Coastal bermudagrass, fescue, and orchardgrass hays at 0, 40, 60, and 80% of the total as-fed substrate, and fiber digestion kinetics were determined in vitro. Kinetics were estimated by the model R = Doe-k(t-L) + U where R is residue remaining at time t, Do is digestible fraction, k is digestion rate constant, L is discrete lag time, and U is indigestible fraction. Parameters of the model were estimated by logarithmic transformation and a direct nonlinear least squares procedure. Corn and wheat starch did not differ in their effect upon lag time of fiber digestion, digestion rate, or potential extent of digestion. Alfalfa had a shorter lag time of fiber digestion (.86 h) than Coastal bermudagrass (3.05 h), but not than orchardgrass or fescue (1.66 and 2.42 h). Orchardgrass differed in fiber digestion rate (.0542h-1) from Coastal bermudagrass (.0698h-1) but not from alfalfa or fescue (.0670 and .0658h-1). The potential extent of fiber digestion was similar for fescue (75.8%) and orchardgrass (76.0%). The potential extent of fibre digestion for alfalfa (50.9%) differed from Coastal bermudagrass (64.3%), and both of these forages differed from fescue or orchardgrass. Addition of starch resulted in a linear increase in lag time of fiber digestion, but digestion rate was not affected. Potential extent of digestion was decreased when starch was added.

Factors affecting fiber digestion in ruminants were evaluated with the use of simple mathematical models. These models were constructed to define the dynamic processes involved so that constraints on fiber digestion may be elucidated. The fraction of fiber that is resistant to digestion and the rate of digestion and passage of potentially fermentable fiber were identified as constraints on fiber digestion in the rumen. Fermentation lag was shown to have no direct effect on fiber digestibility. Fiber that is resistant to fermentation by rumen microbes represents a significant fraction of forage fiber and accumulates in the rumen relative to potentially fermentable fiber. The digestibility of fiber that is potentially fermentable is a function of the rate at which the fiber is digested and its retention time in the rumen. Selective retention of potentially fermentable fiber in the rumen is necessary for the maximization of fiber digestion.