STUDY OF INTERACTIONS BETWEEN FORAGES AND CONCENTRATES IN SHEEP .2. DIGESTIBILITY VARIATION FACTORS

Interactions between forages and concentrates were studied in sheep in the course of 2 successive trials. Results concerning intake (variations in forage/concentrate substitution rates) have been presented in a previous paper (Berge and Dulphy, 1985). In the present study variations in digestibility have been examined for different associations between a forage and a concentrate. Effects of digestive interaction have been evaluated by comparing measured digestibility of the diet to predicted digestibility (prediction based on hay digestibility measured when fed alone and concentrate digestibility calculated by regression) (fig 1). The following variation factors were examined: type and proportion of concentrate, type of forage and level of animal feeding. Seven concentrates based on soyabean hulls, wheat-bran, cereals, beet pulp and lupin (trial I), and barley and beet pulp (trial II) were used. Each diet combining one hay and one concentrate was given ad libitum and then at maintenance level to 12 sheep with 0, 30, 60 and 90% concentrate in the diet (3 animals per diet). Digestibility was measured for organic matter, crude protein, crude fiber and starch. Moreover, some of these diets were offered to fistulated sheep to measure cellulolytic activity, VFA, pH and level of NH3 in the rumen (table I). In 4 situations only (soybean hulls and cereals in diets fed ad libitum; and barley twice, in diets fed at maintenance), digestibility with the same hay/concentrate combination did not increase linearly with proportion of concentrate thus showing digestive interactions (cf figs 2 and 3; table III). At decrease in feeding level by the restriction of allocated feeds, resulted in an increase in diet digestibility, more or less with the type of concentrate (tables IV, V). For the whole diets containing 30, 60 and 90% concentrate (n = 193) the average value of the effects of digestive interactions was practically nil, positive and negative values counterbalancing. The average absolute value of these effects was 1.8, and higher for concentrates based on soyabean hulls, cereals and barley. It increased with starch level and decreased with crude protein level of the diets. Digestive interactions mainly concern crude fiber, ie cell wall content of the feeds. However, the mechanisms involved the latter differed from one situation to another. With diets containing cereals, interactions can be explained by a decrease in cellulolytic activity in the rumen due to the intake of a significant and increasing quantity of starch (table VII). With diets containing soyabean hulls, the interaction can be explained above all by a large increase in ad libitum intake. In case a probable increase in digesta flow, and a decrease in cellulolytic activity, can jointly explain an important decrease in soybean hull digestibility. It is concluded that it is possible to minimise digestive interactions and so the decrease in diet digestibility by restricting feeds offered at maintenance level and by preparing diets with low starch content and sufficiently high crude protein content.