Abstract
A technique of partial feeding through a ruminal fistula (Greenhalgh and Reid, Nature, 214: 744. 1967) was used to assess the relative influence of palatability and digestibility on voluntary intake by sheep of diets diluted with 30 and 50% sawdust. The original technique can be applied, with some inaccuracy, to diets where intake and digestibility are positively correlated, but it cannot be logically applied to diets where intake and digestibility are negatively related. A modified calculation is proposed whereby an expected difference in food intake is calculated for the amount of diet dry matter necessary to achieve the same digestible energy or digestible organic matter intake. This modification, rather than using simply observed differences in intake, allows more accurate assessment of the influence of palatability and is applicable even to diets where intake and digestibility are negatively correlated. By this modified technique the relative involvement of palatability was only 6.7% in the food intake of the 50% sawdust diet in comparison to the 30% sawdust diet.