| Nitrogen fixing trees for fodder production - A Field Manual |
B. W. Norton and M.R Alam
Fodder trees provide a significant natural source of feed for grazing animals in many different parts of the world. The relative contribution to the diet will depend on both the availability and accessibility of tree fodder and the feed preference of the animal species at that particular time of the year. Alternatively, tree fodders are often harvested by farmers to supplement low quality feeds offered to animals raised in housed or confined systems. In this case, feed quality is judged by the farmer and not the animal. There is also increasing interest in the intensive cultivation of fodder trees in agroforestry systems, and the value of fodder trees in these systems depends on troth the amount of fodder produced (yield/ha or yield/tree) and the relative intake and quality of tree fodder and understory forage. The establishment of such agroforestry systems requires prior determination of the tree species best adapted to the prescribed environment and the potential nutritive value of the chosen species.
Nutritive value may be broadly defined as the ability of a feed to provide the nutrients required by an animal for maintenance, growth and reproduction. It is a function of both voluntary feed intake (FI) and the efficiency of extraction of nutrients from the feed (digestibility, D). In the absence of anti-nutritive factors, forage intake is often positively correlated with D, and the intake of digestible nutrients (nutritive value = FI x D) is directly related to animal production (live-weight gain or milk production). It is for this reason that techniques have been developed which purport to predict D (in vitro and in sacco digestibility), and these values used as an index of forage quality. Although these techniques have proved valuable for grass forages, there is little evidence to support the use of these techniques for either forage legumes or tree fodders (Jones and Wilson 1987). The reason is that, unlike grasses, there is often little correlation between FI and D in animals consuming tree fodders, and hence estimates of D are not necessarily indicative of digestible nutrient intake. The major factors which might affect nutritive value are therefore voluntary feed intake, nutrient content of the feed and the extent to which nutrients are extracted from the ingested feed.
Chemical composition and its interpretation
Chemical composition alone, as measured by the proximate and elemental analysis systems, is an inadequate indicator of nutritive value. These measurements take no account of either the form or availability of nutrients, and, at best, may provide information on potential nutrient content. Analyses based on detergent extraction are more useful since plant dry matter is separated into a completely digestible fraction (neutral detergent solubles, NDS) representing cell contents, and a partially digestible fraction (neutral detergent fiber, NDF) representing plant cell walls. There is now an extensive literature on the chemical composition of fodder trees (Skerman et al. 1988, Norton 1994a, Devendra 1995) from which some general guidelines can be drawn (Table 1). The crude protein (CP) content, N content x 6.25, of fodder tree leaves is usually high (150-250 g/kg) and exceeds that required for the maintenance of fermentation in the rumen (65-85 g/kg). Unlike grasses, the protein content of tree leaves remains relatively constant throughout the year. However the high CP content of fodder trees leaves may not confer any direct nutritional benefit to the animal unless these proteins are protected against ruminal degradation. The presence of condensed tannins in some fodder tree leaves affords some protection against ruminal degradation, and thereby delivers significant amounts of plant protein for absorption in the intestinal tract. Macro- (Na, K, Mg, Ca, S) and micro element (Cu. Mn, Zn, Co, I, Se) analysis provides values which may be compared with animal requirements. However, while values less than predicted requirements are indicative of deficiency, values greater than requirement are not necessarily indicative of sufficiency. There are few reports of elemental deficiencies in fodder tree leaves, and because tree leaves seldom form all of the diet, deficiencies are likely to be rare under practical feeding conditions. The following feed composition measurements are therefore considered most useful for describing the potential nutritive value of tree leaves: dry matter (DM), organic matter (DM-ash), crude protein and NDF.
The presence of toxic and anti-nutritive factors
Low palatability of tree leaves may be related to either physical (hairiness, steminess, etc) or chemical feeding deterrents. Many fodder trees contain saponins, alkaloids or polyphenolic compounds (Norton 1994b), and the detection and analysis of these substances often requires sophisticated analytical techniques not commonly available in most laboratories. The potential toxicity of tree leaves may be determined by feeding in a diet to laboratory rats. However toxicity in rats may not always be seen in ruminants which may degrade potential toxins in the rumen. It has also been demonstrated that the effects of some potentially toxic substances (e.g. mimosine from Leucaena) may be considerably reduced where leaves form less than 30% of the diet. It has also been shown that drying or wilting may alter the potency of some toxins. The detection and measurement of condensed tannins is a relatively simple procedure, and should be included in routine analytical measurements on fodder trees where possible.
Predictive techniques for digestibility
The measurement of digestibility is useful where it provides a prediction of the intake of digestible nutrients. Where possible, feeding trials to measure both voluntary feed intake and in vivo digestibility should be undertaken. However, as discussed later, for fodder tree leaves such measurements have a limited value. There is evidence that factors other than the rate of digestion in the rumen determine the voluntary intake of tree foliage by ruminants. Low leaf intakes may be associated with high leaf digestibility if the leaves contain appetite depressants such as alkaloids and tannins. Alternatively high feed intakes associated with low leaf digestibilities may be the result of the rapid passage of small leaflets from the rumen. Many leguminous fodder trees have pinnate leaves containing small leaflets. Lowry (1989) has observed that sheep consumed significantly more fallen Albizia lebbeck leaf (DM digestibility 43%) than fresh leaf (DM digestibility 64%), supporting the view that the measurement of digestibility alone is of limited value for the prediction of nutritive value. Although in vitro and in sacco (incubation in ruminal fluid) assays are commonly used to rank feeds in terms of digestibility, such assays are of little value for ranking fodder trees in relation to nutritive value. The continued use of such assay demands that a relationship be established between these assays and feed nutritive value.
Feeding trials to determine nutritive value Tree fodders are most commonly used as supplements to low quality forages rather than sole feeds for ruminants. This suggests that the most important information about the nutritive value of tree fodders is the level of supplementation which optimizes nutrient intake. This level will vary with the quality of tree fodders and of basal diet, and will vary with different plant parts and fodder tree species. The following technique has been adopted at the University of Queensland to provide information on: a) a comparison between low levels of fodder tree leaf supplementation with urea/molasses, b) the level of supplementation which maximizes nutrient intake, and c) the nutritive value of fodder tree leaves as the sole feed source. These results are obtained by offering a low quality diet (barley straw) ad libitum to 3 groups of animals. Fodder tree leaves were then provided at rates of O. 1% and 2% of liveweight to each group. A fourth group was offered tree leaves ad libitum without barley straw. The group without leaf supplements was given a mixture of urea and molasses (100g containing 30% urea), this supplement providing about the same amount of N and digestible energy as the lowest level of tree leaf supplementation. Table 2 shows some results obtained from these trials (Norton et. al. 1992), and clearly demonstrates the relative merits of the different fodder trees as supplements for goats. As expected, digestible DM intake (nutritive value) was closely related to animal production (liveweight gain). It is significant that for both Leucaena and Sesbania, there appears to be no benefit in levels of supplementation greater than about 33%, but for Albizia, net nutritive value decreased as levels increased above 30%. In this case, Albizia was of low nutritive value when fed as the sole feed, and its nutritive value was maximized by feeding at low levels (27%) in the diet. It is also clear from Table 2 that when tree leaves are fed as the sole source of feed, digestible DM intake may be a poor indicator of nutritive value. This information is useful for making recommendations on optimum levels of fodder tree leaf use in cut-and carry systems and in agroforestry grazing systems where intermittent use is made of fodder tree reserves.
Grazing trials and animal productivity
Although the establishment and implementation of grazing trials to evaluate the productive potential of fodder trees is a long term and expensive investment, these trials provide definitive information on the value of fodder trees in animal production systems. Introduced fodder trees may be used in many different ways in grazing systems. They may act as an autumn or winter reserve for strategic grazing, as a sole source of feed for the intensive finishing of stock or as an integrated forage source which is continuously grazed by stock. Prior to the commencement of such trials, the objectives of the experiment must be clearly defined, and the number of animals and areas of land used be sufficiently large to permit meaningful statistical analysis of the data. Appropriate control treatments must also be included to allow comparison with existing systems under the same environmental conditions. A minimum period of one year is required to permit assessment of seasonal effects on tree and animal productivity. The fodder tree banks established should be mature enough to withstand the stocking pressure applied, and in some circumstances it may take up to 3 years before tree banks can be grazed. Stocking rate is the major determinant of animal productivity from pasture, and an appropriate range of stocking rates should be applied to test the potential productivity of the system. It is not possible to be prescriptive about the design of such experiments, since different environments and systems of production will impose particular constraints on experimental design and interpretation.