With 2 figures in the text)We describe food selection by roe deer (Cuprealus capreolus) in relation to the food quality of the plants available (the concentrations of fibres, sugars, crude protein, and of phenolics and terpenes). Seven tame roe deer feeding in an oak-beech woodland edge used the majority (SO-94%) of the plant species available: they were therefore generalist feeders. However, they preferred only a small number of plant species in the different seasons: ivy (Hederu helix) in winter and autumn, dogwood (Cornus spp.) in summer, hornbeam (Curpinus befulus), hawthorn (Crutuegus spp.) and bluebells (Hyacinthoides non-scriptu) in spring. No preference or avoidance could be demonstrated for 74-85% of the available species, but 7-12 species were avoided according to the season. Though the avoided species were sometimes common, they never comprised an important part of the diet. In contrast, the preferred species made up only 4-13% of the available green matter but composed a large part of the diet (22-49%). Although generalists, the roe deer were therefore also highly selective feeders. The use of the different plant species was influenced by their availability, but to a different extent in the different seasons. Use of the plant species correlated negatively with the fibre content, as predicted from the digestive morphology of this concentrate selector. Preference/avoidance was related to the concentrations of soluble sugars, but not the crude protein content. Contrary to expectation, the roe deer preferred plants with a high concentration of protein-binding phenolics, suggesting that this deer has specific mechanisms for de-activating these compounds. Among the nutritional consequences of the feeding strategy of roe were a 17-35% higher soluble sugar content of the diet compared with the available green plant tissues, and an even greater increase in the phenolics. The crude protein contents and the dry matter digestibility of the diets tended to be slightly higher, and the fibre contents were lower, than those of the available green plant tissues.

Organic and mineral components have been analysed on 18 dicotyledonous species of permanent grassland communities. Multivariate analysis revealed an opposition between a macromolecular pool (cellulose and lignin) and a nutrient pool (potassium and phosphorus, nitrogen). The first pole is characterized by Tragopogon pratensis L. (Compositae) and Galium mollugo L. (Rubiaceae), the second by Anthriscus sylvestris (L.) Hoffm. (Umbelliferae) and Geranium sylvaticum L. (Geraniaceae). The most thermo-helio-xerophilous species (Knautia arvensis (L.) Coult and Tragopogon pratensis) belong to the first group, whereas the second group includes the cool-environment species (like Alchemilla xanthochlora Roth.), which reveals an ecophysiological determinism. Moreover, correlations between the biochemical and ecological structure, and other interrelated parameters, including sclerophylly, have been shown. Compared to lignous and sempervirent plant communities, the grassland species have lower phenolic and lignin contents but higher glucide and nutrient (potassium and phosphorus) contents, which is in accordance with their food value. For this herbaceous community, the C/N ratio indicates the same general equilibrium between cell-wall macromolecules and nitrogen than for two other plant communities (mainly ligneous) previously studied, with some differences revealing distinct trends from the same general metabolic sheme.Key words: dicotyledonous plant communities, biochemical organization, C/N ratio, ecophysiology.

No abstract

cellulolyse / cellulase / dégradation / inhibition / phénols Summary—A study of the influence of different plant species from natural grasslands on enzymatic hydrolysis of cell-wall polysaccharides. A simple, rapid and repetitive method for the estimation of the potential of herbage to inhibit enzymatic activity is described. The enzymatic degradabil ity of the cell walls (the factor determining the organic matter digestibility) was quantified by the phenolic compounds liberated by cellulase action of their ester linkages with cell-wall carbohydrates. Inhibition of this action by some plant compounds solubilised during refluxing in a phosphocitric buffer solution (pH 4.6) was expressed as an index of potential negative action (IANP). This index expresses the action of inhibitory compounds in plant extract on the cellulolytic activity of a fungal enzyme as compared with the activity of this enzyme measured without plant inhibitors. This is useful in the prediction of the potential negative action of plant species on the quality of the herbage association of which they are a part. The determination of IANP in many natural grassland species shows important differences between grass, legumes and other dicotyledonous plants and also within these botanical groups. This index can be useful for the determination of an acceptable proportion of individual plant species in herbages.

No abstract