Assessing the nutritional quality of stored grain legume fodders : Correlations among farmers’ perceptions, sheep preferences, leaf-stem ratios and laboratory analyses

Crop residues have the potential to alleviate annual feed shortages and nutrient deficiencies experienced in the dry season in the savanna zones of West Africa. Farmers in West Africa especially value the residues of grain legumes, also known as grain legume fodders (GLFs), as animal feed. In this study, therefore, we assessed the nutritional quality of GLFs as affected by storage conditions using four different methods: farmers’ perception score (FPS), sheep preference score (SPS), leaf-to-stem ratio (LSR), and laboratory analysis of organic matter digestibility (OMD), crude protein content, neutral detergent fibre (NDF) and acid detergent fibre (ADF). We also determined correlations among these variables. The fodder of cowpea, groundnut and soybean were stored separately in three locations (rooftop, room and treefork) and with two packaging types (polythene sacks or tied with ropes) for 60, 90 and 120 days. FPS was determined by scoring the perceived quality of GLFs on a scale of 1–10 (1 = bad and 10 = good) based on physical characteristics by a group of farmers. SPS was assessed by a cafeteria feeding trial based on dry matter intake of GLFs by a flock of 12 sheep per village during a 14 hr period. LSR was determined based on the mass of the botanical fractions, i.e. leaf (leaf blade only) and stem (stem and petioles) of 200 g samples separated carefully by the hand. Laboratory analysis was done by near-infrared spectroscopy (NIRS). Results showed that all quality assessment methods successfully discriminated GLF quality differences among crops. Only farmers and sheep could distinguish quality differences among all storage conditions and packing types, whereas laboratory analyses methods could not. These findings could be due to the fact that farmers use LSR to evaluate feed quality, though colour, texture and smell of the fodder could also contribute. We also found significant correlations (ranging from 0.35 to 0.88) between all the quality assessment methods across all treatments. There were few within crop correlations between the fodder quality assessment methods, i.e. only FPS and LSR for groundnut and cowpea, FPS and CP for groundnut and all laboratory analyses parameters among each other for all crops. Hence, the differences among crops were the important determinants of the correlations. From this study, we conclude that farmers have experience and knowledge about nutritional quality of feed and livestock preference for feed. Development programmes and projects could benefit from using such knowledge when formulating and implementing interventions