Tropical resources : availability and feed value

Abstract

Availability and sources of variation in the feed value of tropical plant resources and agro-industry by-products were studied. There is a large diversity of feed resources (grass, legume, fodder trees, grains, tubers, co-culture, by-products of agro-industries) reflecting the great diversity of more or less intensified farming systems that includes feeding the different crops to livestock at variable levels. Feeds may be identical to those (soybean, corn) used in temperate zones, especially in modern farms. They can also be classified by their botanical origin (sorghum, tubers, fodder, and fodder trees), their composition (frequent presence of secondary metabolites) and treatment technology (agro-products from artisanal food) that they have suffered. There is great variability between and within feed value of resources. Whatever the animal species, high energy value resources, similar or close to temperate counterparts, are available. In contrast, the many protein-resources have values that do not « compete » with soybeans. Specific strategies to use certain resources are necessary because of the presence of secondary metabolites with antinutritional activities. Intra resource, management (fodder age, feeding strategy...) and the technology used (elimination of antinutritional factors, method of oil extraction from oilseeds, processing of cereal grains) induces variabilities that affect both the energy and protein values and that are often higher than those resulting from genetic origins. High feed value resources are available in the tropics provided that one applies the proper management and chooses the right animal species for their valorsiation. The accumulated information on the feed value of tropical resources is abundant. Access to data is not always easy because the publication strategies are often focused on a regional scale. Information can be partial (some criteria of chemical composition). The results can be strongly linked to the context in which they were produced, which limits their generalisation. An important step for the future is the collection of these data with the prospect of generating such general laws. This would be possible through the current FV table project (AFZ-INRA-CIRAD).