Previous Page Table of Contents Next Page

Practical production of protein for food animals - Stephen A Chadd, W. Paul Davies and Jason M Koivisto

Stephen A Chadd, W. Paul Davies and Jason M Koivisto
Royal Agricultural College,
Cirencester, Gloucestershire,


Demand for meat could increase by 58 percent between 1995 and 2020 according to IMPACT food model predictions of the International Food Policy Research Institute (IFPRI). Poultry meat demand might increase by 85 percent; beef by 50 percent and pigmeat by 45 percent over this time period. IFPRI also predict that 97.5 percent of the population increase up to 2020 will be in the developing world, representing at this time 84 percent of global society. Income growth; urbanization; changes in lifestyles and food preferences in addition to continuing population growth could double the demand for meat in the developing world up to 2020. Other drivers for change in the agri-food sector include advances in technology; regulatory requirements and institutional pressures; environmental considerations; globalization influences; competition and political intervention. All of these factors, to a greater or lesser extent, will impact on the so-called ‘livestock revolution’. Future feed sources and supply to support the substantial growth in livestock production, as well as the approaches to livestock husbandry, are a continuing cause for concern. Protein availability and supply is a particular concern, especially in the light of meat and bonemeal restrictions, the adoption of genetically modified crops, dioxin residues in fishmeal and increasing pressures on fisheries policy. Sources of protein are reviewed, including by-products of the food industry, oilseeds and arable and forage legumes. Alternative, and currently less common, plant protein sources are assessed.

More information is required on less popular protein plants to clearly identify the reasons for relatively low adoption. A much greater emphasis is recommended for improving plant protein supply in marginal growing environments. The increasing importance of both technical and safety aspects of protein product quality is stressed. Better technology transfer and small farmer support is considered essential for encouraging further protein crop advances.

More research is recommended in the short and medium term on agronomy and the further development of alternative, and novel protein supply cropping. More focussed support for longer-term strategies of crop improvement, through both breeding advances and genetic manipulation, is urged. More meaningful and greater co-operation is advocated between policy-makers; the feed industry; farmers and researchers to better deliver the future protein supply potential.


Domestic animals make critical and valued contributions to society and human existence throughout the world, and play a key role in agriculture. Livestock products account for an estimated 30 percent of the total global value of food and agriculture, and approximately 19 percent of the value of global food production (Heap, 1998). Products from food animals provide over 33 percent of protein consumed in human diets globally and about 16 percent of food energy (Martin, 2001). Non-foods such as wool, hides, bones and dung for fuel are also valuable commodities. Animal manures make a very important contribution to soil fertility, particularly to productivity in the developing world. Animals also provide important power for cultivations and transport in some societies - and globally represent considerable value, equity and insurance.

Protein is an essential key ingredient of animal feeds. It is absolutely necessary for animal growth, body maintenance, the production of young and the output of such products as milk, eggs and wool.

Approximately 11 percent of the global land mass is cultivated and about 26 percent is permanent pastureland, with 31 percent in forest. In traditional low output farming systems the protein supply can be met from plants and crops grown locally. Higher output animal production is now increasingly important for commercial livestock and mixed farm viability, and nutrition (particularly protein) requirements have become much more demanding. High performing animals need higher quality feed and, except for extensive sheep and beef systems, imports of quality protein and energy are now the norm in the form of compound or straight feeds. Some 800 million tonnes of compounded animal feeds are now produced annually worldwide (IFIF, personal communication, 2002).

In considering protein budgeting and use, a clear distinction has to be made, of course, between requirements for monogastrics such as pigs and poultry, and ruminants such as cattle and sheep, where bacteria in the rumen significantly influence protein synthesis and absorbable amino acids.

Particular concerns, and justification for the current United Nations Food and Agriculture Organization consultation on ‘Protein Sources for the Animal Feed Industry’, are being generated by increasing safety considerations (either real or perceived) in key protein sources. These include the prospects for genetic modification (GM) in major crops and its public perception; potential human health risks from meat and bonemeal - highlighted by the Bovine Spongiform Encephalopathy (BSE) crisis, and dioxin residues in fishmeal.

Safety, and potential health hazards are not the only issues challenging farmers and animal feed manufacturers in a rapidly changing agri-food industry and trading environment - but they remain critically important for consumer confidence in particular and for government reaction.


As always, the future remains uncertain - and mostly unpredictable. It is quite clear, however, that the agri-food sector is experiencing accelerating rates of multi-dimensional change. No one factor is changing on its own. These drivers for change (Davies and Turner, 2002) include:

These on-going influences in themselves can be multi-dimensional, as in the following examples - and the impact on the agri-food industry is considerable.


On the demand side, for example, consumption will involve considerations of demographics and population structures; cultural and religious issues; economic status and disposable income; aspirations; concerns and lifestyle. Consumers in the United Kingdom, for example, are ‘more demanding’, ‘better informed’ and ‘more discerning’ - and are demanding more ‘transparency’, ‘traceability’ and ‘assurance’ in the food chain. A recent Food Standards Agency (FSA, 2001) survey highlighted the following significant influences on food purchases in the United Kingdom (when not prompted) of:

When specifically asked to rank various given factors on the basis of being ‘very or quite important’ - ‘health’, ‘taste’ and ‘food safety’ scored highly.

As a probable result of the many recent health and safety crises in the United Kingdom food chain, many respondents (FSA, 2001) are concerned about the methods of food production - 16 percent of men and 27 percent of women said they were concerned about how animals are treated and raised, confirming the significance of ‘animal welfare’ issues in livestock production in the United Kingdom. Since the BSE crisis, quality assurance schemes in relation to meat products in particular, have become important in the United Kingdom - and are becoming increasingly important in international trade (Baines and Davies, 1998).

‘Understanding the consumer’ is likely to become increasingly important as a major driver in livestock system development, including feed supply. Although in other countries, where the market for livestock products is still developing, there may currently be less concern for ‘quality’ issues than quantity or affordable prices - cardinal quality concerns (such as consumer safety linked to production approaches) are still likely to become increasingly significant. Understanding these developing markets, and the developing consumer behaviour and requirements, will be the key to greater understanding of livestock production chains.


Of the other major drivers, the need to improve competitiveness and production efficiency is becoming increasingly significant, for both domestic markets and international market positioning in particular. Benchmarking performance is increasingly important in this respect, as is the need to remove unnecessary costs in the supply chain. The costs of protein, and diet formulation, for livestock production in relation to performance is a key issue in this regard. Least cost programming to determine the value of each protein source will remain critical.


Technical advances, through research and development, will continue to make a particular impact, and beneficial advances need to be carefully explained and effectively communicated to all involved in the food chain - and to consumers in particular.

Transgenic (GM) crops are currently grown on over 45 million hectares globally (James, 2001). The main GM crop being planted on 26 million hectares, 58 percent of the transgenic area, is herbicide tolerant soybean grown mainly in the United States, Argentina and Canada but also on smaller areas in Mexico, Uruguay and Romania. In second place is transgenic corn, planted on 10.3 million hectares followed by GM cotton on 5.3 million hectares and GM oilseed rape on 2.8 million hectares (James, 2001). The impact on the animal feed industry and on agriculture in some areas is already significant. Hesitancy in the European Union (EU) over the adoption of GM crops and consumer concerns could have wide-ranging effects on farming approaches, livestock feeding and competitiveness.

There is little doubt that genetic modification of crops and livestock will potentially, have an increasing impact - but so probably will advances from conventional breeding. The influence of new manufacturing processes (and innovation) in the animal feed industry should not be underestimated either.

Progress will undoubtedly be made on new protein sources, and different attributes of protein feeds. A greater focus on protein quality rather than simply on crude protein as a measurement of protein supply would also seem likely.

What might the impacts be, for example, of (in a European context initially) higher digestibility grasses, exploitation of the ‘stay green’ gene and proteolysis prevention during ensiling? Or the impact of improved ‘naked oats’ as a candidate cereal for improving protein supply? In the longer term, what will be the outcome of plant breeding efforts to increase protein values in barley - or to widen the climatic tolerances of exotic protein crops such as soybean? Some of the technical difficulties may presently be substantial, but how soon can they be resolved?


Greater concerns globally for a more sustainable agriculture, with a lower environmental impact, will have an increasing influence on farming systems and future approaches to food production. The real and potential influences of intensive livestock systems on environmental pollution is regarded as a serious issue in many countries - leading to further encouragement of less intensive approaches. Agri-environment schemes, to promote better countryside stewardship are becoming particularly important in Europe (Curry, 2001). Environmental perspectives will also modify future approaches, both directly and indirectly, to protein cropping. The use of break crops (utilizing peas or beans) in an arable rotation or a legume-rich (e.g. white clover) grass ley in cropping is being increasingly advocated as part of an integrated crop management approach in Europe. A much better understanding of the fate of nitrogen in mixed rotational farming systems is, however, required (Warman et al., 1997). The value of rotations is being reassessed and is gaining political ground. Greater usage of prospective spring-sown legumes, in an attempt to boost home-grown supply, will increase the area of bare ground over winter and in temperate systems, the potential for nitrogen leaching. The debate concerning the optimization of nitrogen usage on crops such as cereals and grassland, and political threats of a ‘nitrogen tax’ to protect the environment, are also likely to continue (at least in Europe).

The extent to which new approaches to inter-cropping and companion cropping with legumes can mitigate against deleterious prospects in commercial temperate cropping in north-west Europe, remains to be seen.

From an animal feed perspective, future supplies of fishmeal may also be affected by concerns for over-fishing and a more sustainable fisheries policy.


Political support, or otherwise, for changes to agricultural systems and food production will continue to have a major influence on protein supply issues and the global realities of the animal feed business. For example, continuing encouragement of ‘home-produced protein cropping’, integrated crop management, organic farming and non-adoption of GM crops, will have significant influence in the European Union.

United Kingdom Government support for more research on soybeans, sunflowers and lupins or alternative protein crops could result in significant medium-term advances in production. Whereas, government encouragement and support for soybean production in North and South America will continue to have a major impact on global supply.


Global demand for cereals could increase by 39 percent between 1995 and 2020 (to 2 466 million tonnes), meat demand by 58 percent (to 313 million tonnes), and roots and tubers by 37 percent (to 864 million tonnes) according to IMPACT (International Model for Policy Analysis of Commodities and Trade) predictions. This global food model of the International Food Policy Research Institute (IFPRI), also predicts that 97.5 percent of the population increase between 1995 and 2020 will occur in the developing world. This rise to a possible 6.3 billion people in these countries by 2020, would represents 84 percent of the global population (Pinstrup-Andersen et al., 1999).

Most of the future world food demand will therefore occur in developing countries. An estimated 85 percent of the increase in demand for cereals (690 million tonnes) and meat (115 million tonnes) between 1995 and 2020 could occur in the developing world.

Up to 2020, demand driven meat consumption in the developing world will grow three times faster than in the developed world. Total demand for meat will double. To supply such a massive increase in livestock production, the cereal grain supply may need to double, and the demand for maize in particular will be considerable.

IFPRI predict that by 2020, up to 60 percent of the cereal demand in developing countries may have to be imported and meat imports increased significantly (to 6.6 million tonnes) (Pinstrup-Andersen et al., 1999).

By 2020 however, due to the population increases, an average person in a developing country will consume less than 50 percent of the cereals consumed by a developed-country person, and only about 34 percent of the meat products.

The IMPACT model projects that between 1995 and 2020, poultry meat demand will increase by more than 85 percent; beef by 80 percent and pigmeat by 45 percent.

Food insecurity and malnutrition will continue to be a significant and serious problem for the foreseeable future. Sub-Saharan Africa and South Asia in particular will remain as problem regions for food insecurity up to 2020 (Pinstrup-Andersen et al., 1999).


Foods of animal origin provide about one-sixth of human food energy consumed globally, and one-third of the protein (Martin, 2001). Rapidly growing demand for livestock products worldwide is brought about mainly by human population increases and growing income, but also by changing lifestyles, food preference and urbanization (Conway, 1998; Avery, 1998). Land use and human population pressures are leading to intensification and expansion in many livestock production systems. Like the Green Revolution, the ’Livestock Revolution’ defined by Delgado et al. (2001), involves the large-scale participation by developing countries in farming transformations that had previously occurred mostly in temperate zones of the developed world. These rapidly expanding livestock sectors are inevitably exerting increased pressures on natural resources. Technologies are needed to increase the efficiency of feed conversion (thereby reducing inputs and nutrient losses), and to develop more sustainable production systems and product-use.


The quality of human nutrition is inextricably linked to the quality of the livestock products consumed which, in turn, is significantly influenced by the nature of the raw materials (protein sources) eaten by the animal. There are those who feel that grain production should replace livestock production on land used for grazing, and that it is poor food economics to feed grain to animals. It could be argued, however, that much of the land used to produce livestock is not suitable for grain or alternative crop production anyway - and good, consistent meat quality can come from grain-fed animals. The link between affluence in societies and the demand for higher dietary levels of high-quality protein has been made by Avery (1998). Some examples provided include Japan, whose population have in recent years, increased their protein intake (mainly of fish) from 20 to 60 grams per day. Such consumption patterns are being emulated in Taiwan and South Korea. Poultry production increases in Thailand also reflect the importance of, and desire for, elevated levels of meat consumption (and export).

In contrast, and particularly noticeable in developed countries in northern latitudes, is the dietary shift away from animal to plant sources of protein. This being perceived as the healthier option (Millward, 1999). In addition to the move by some towards vegetarianism, is the desire by others for white meat in preference to red. It has been calculated that approximately 70 percent of the total animal protein eaten by humans is provided by ruminant animals, (Minson, 1997) and that 35 percent of all protein consumed is derived from animals. The debate continues about the nutritional adequacy of plant-based diets versus meat (Sanders, 1999). In addition to meat as a vital source of dietary protein for humans, milk and eggs are also standards against which the adequacy of other protein sources can be measured in terms of the provision of essential amino acids for body protein synthesis. Gill (1999) and Rosegrant et al. (1999) emphasize the significant variation there is in the proportion of meat consumed in national diets across the world - again linked to population growth, income and the degree of urbanization. Much of the debate in the West continues to focus on ethical and environmental issues, relating to the production methods employed in livestock systems.


A comprehensive classification of types of livestock production systems has been produced by Sere and Steinfeld (1996) and further discussed by Gill (1999). The demand for livestock products and commodities in different parts of the world (as previously discussed) has influenced the characteristics of livestock systems, including the types of feeds used and observable trends. Livestock make an important contribution to most economies. The criteria used by Sere and Steinfeld (1996) to characterize livestock systems included regional differences, quantitative estimates of the importance of each system globally, human population dynamics, livestock numbers and outputs. Three main patterns are identified which effectively describe the nature and diversity of global systems, namely: grazing, mixed farming (some feed from crop residues and by-products produced on the farm) and so-called industrial or ‘landless’ systems. In excess of 65 percent of the world’s cattle population and small ruminants are located in the developing world, but factors such as culture, climate and economics will determine how much variation there is between countries and regions (Gill, 1999).

The landless intensive production systems, mainly monogastric, tend to predominate in developed countries and are responsible for producing more than half of the total meat production. The Organisation for Economic Co-operation and Development (OECD) member countries account for a significant proportion of total pork and poultry production globally in landless systems. Asia is next in importance for pork production and in the case of poultry, Central and South America. Such systems are very reliant on imported feed raw materials, capital intensive and associated with concerns over environmental pollution.

Landless ruminant production systems are focussed in a few regions of the world. Eastern Europe and the Commonwealth of Independent States (CIS) seem to be the preferred locations for most landless cattle systems, and sheep farmed in this way are found mainly in Western Asia and North Africa. Typical examples are also the large-scale feedlots of cattle in the United States. Under such systems the ruminants are fed essentially as a monogastric, with high levels of concentrates and cereal-based diets. Such systems are associated with substantial economies of scale. In respect of trend, very intensive types of farming operations can be expected to decline in importance in the European Union as production becomes more extensive, and in response to policies which promote protection of the environment and the gradual removal of agricultural support.

To put the systems in perspective in terms of output of meat, only 9.3 percent of the total is produced in grassland-based systems, compared with 37 percent in landless and 5.3 percent for mixed farming operations (Sere and Steinfeld, 1996). The fastest growing meat production systems are the intensive landless ones with growth rates (percent per annum) of 4.3, 2.2 and 0.7 for landless, mixed and grassland systems respectively. Along with landless monogastric systems, the importance of mixed systems as suppliers of livestock products is expected to continue to grow in the future.


An example of the merits of crop-animal (mixed) systems is provided in Asian agricultural practice in which livestock have a multi-purpose role (Devendra and Thomas, 2002). Although animals are considered secondary in importance to that of crop production, the complementarity of the two systems can readily be seen particularly from a systems’ sustainability viewpoint. At a local level, most of the projected future demands for ruminant meat and milk are expected to be met from the improved productivity of livestock in these mixed farming systems. Further research would be welcome on the more effective use of crops as potential animal feed and, in particular, alternative protein sources. Annual and perennial crops are grown, including tree species, and these are closely integrated with the livestock sectors. Thomas and colleagues (2002) elaborate on the factors which represent potential constraints to Asian livestock productivity - including inadequate genetic resources, feed sourcing problems, health and disease and in places, poor infrastructure. Perhaps the greatest challenge confronting the livestock system is to increase the availability of animal feeds, both in respect of quantity and quality (particularly protein content). Feed deficits exist throughout South Asia as a whole, with significant regional differences. Feed protein sources and their use tend to be prioritised according to their perceived quality and the animal productivity level desired. Although some nutritional improvement to roughages, such as ammonia treatment, have been tried on farms, the more effective way to raise production standards has been by supplementation with cottonseed and oilseed cakes. However, the cost is often prohibitive for many farmers and the performance results disappointing and variable (Devendra and Sevilla, 2002).


One of the major contributors to the cost of production in livestock farming, particularly pigs and poultry, is the price of protein per unit weight of animal feed. For the European Union, with implementation of the ban on the use of meat and bonemeal together with the predicted demise of fishmeal, there is the realisation that a bigger market will be created for alternative protein feed sources. The biological value of meat, bone and fish meals in terms of their recognised amino acid profile, will be difficult to substitute. The ban substantially reduces the European Union’s feed protein self-sufficiency. The experience of the United Kingdom is characteristic of most European Union countries. The increasing reliance on imported proteins (Merry et al., 2001) also increases farmers’ and compounders’ exposure to the price fluctuations, currency movements, supply shortages and surpluses associated with the main protein source - soybean. Globally, there can and will be imbalances between the production and availability of suitable livestock protein sources. Any resulting increased competition on the world market could, for example, give a competitive advantage to countries which can readily produce a feed such as soya (ENTEC, 1998). The cost of importing protein concentrates can be significantly influenced by supply and demand and other market forces, all of which can be extremely variable.

For any livestock farmer, reducing costs whilst maintaining a desirable level of output, is an objective which not all are able to realise. The breeding and development of new protein crops, or the enhancement of the nutritional value of indigenous crops, offer the potential to increase competitive advantage. The efficiency with which protein nitrogen in the animal diet is converted into products varies according to species (ruminant versus non-ruminant), the stage of the production cycle and quality (in amino acid profile terms) of the feed protein being offered. In the landless monogastric livestock systems referred to earlier, larger intensive units exercise great care and precision in diet formulations, which amongst other things, allows them to incorporate synthetic amino acids, for example. The main benefits of the creation of an 'ideal protein' are twofold. Firstly, matching supply with the animal’s nutrient requirement reduces the chance of environmental pollution and secondly, it can make economic sense.

Efforts are already being made in Europe to rectify any shortfall in traditional protein feed sources. This includes providing incentives for farmers within the European Union to expand their plantings of soybeans, field peas and beans to meet the increased vegetable protein requirements. A possible problem with this policy may be that increased production of such crops could be counter to the Agenda 2000 reforms of the Common Agricultural Policy. Under Agenda 2000, the European Union is harmonising area payments for oilseeds and grains over a three-year period. It is also very uncertain how much soya the European Union could grow, particularly in northern latitudes. In considering alternative protein sources, it is important that governments and others appreciate the vital part that pastures and forage can play in supplying ruminants with their protein needs (Merry et al., 2001). The purpose of recent research in Australia by Robinson and Singh (2001) was to evaluate alternative protein sources for laying hens. There was a concern over increasing soybean imports, and the realization that cultivation of indigenous legumes (mung bean, chickpea and cowpea) and canola could reverse this trend and enable an increased level of self-sufficiency. Some legumes were found to be very well suited to sub- tropical regions and showed considerable promise as competitive sources of protein for livestock - in this case for poultry production.

In parts of the developing world there is not perhaps the same luxury of being able to select alternative quality protein sources for livestock production as there is elsewhere. This may be for economic reasons. Teferedegne (2000) reports research carried out to enhance the productivity of ruminants in sub-Saharan Africa. The primary feed sources were roughage with poor digestibility. Rather than supplementation with oil seed cake proteins, the nutritional value of local tree legumes (Enterolobium, Samanea and Acacia species) were investigated to substantiate claims of the beneficial effect on increasing nutrient intake (including nitrogen) and improving feed conversion efficiency and thereby animal performance. The protein content of forage tree legume leaves is usually high (150-300 g/kg) compared with that of hay and crop residues (30-100 g/kg). Although the presence of anti-nutritional substances in tree-based forage can potentially inhibit intakes and utilization, such legume sources appear to have a positive effect on rumen micro-organism function. These studies confirm the potential benefits to livestock production that could be achieved through the introduction of fodder trees and shrubs. Further research could and should also explore ways of improving the quality of crop residues through breeding or chemical treatment. Knowledge at farm level should also be provided to show how to incorporate non-conventional protein substitute feeds into animal diets.


There are a wide range of factors and influences that need to be considered when assessing the future of crop production.


There are many possible sources of plant protein for livestock rations. These include oilseeds, by-products of food production, arable and forage legumes.


There are many examples of high quality plant protein sources available from by-products of food or drink production (Crawshaw, 2001). Two important examples illustrate the potential of using these materials for livestock feeding. Brewers’ grain and maize gluten meal (MGM) are two of the most common sources of by-product protein.

Brewers’ grain is a term that applies to a broader group of spent grain products, including the non-alcoholic malting industry. Crawshaw (2001) found that brewers’ grain can be highly variable in terms of its crude protein content, varying between 170 and 320 g/kg, with a mean value of 240 g/kg. With such variability in protein content, it is very important for producers to check the protein value of specific batches of grain. Brewers’ grain can be fed fresh to stock, or ensiled for feeding at a later date. It can provide an excellent source of protein in high roughage diets with inclusion rates as high as 8 kg dry matter (DM)/d. It was found in United Kingdom studies (CEDAR, 1995) that brewers’ grain increased milk yield and milk protein, but reduced milk fat concentration.

Maize gluten meal is a product of the maize fractionation process which involves extraction of starch, germ and bran from the grain. MGM tends to have 600 to 700 g/kg crude protein, making it one of the richest potential protein sources. However, MGM is deficient in lysine with 17 g/kg, compared with soya meal which has 62 g/kg. This lack of lysine makes MGM a poor source of protein for pigs without lysine supplementation. MGM is however a good choice of protein for egg producing poultry, because of the relatively high methionine content of 28 g/kg compared with 14 g/kg for soya (Crawshaw, 2001). Due to the relatively high proportion of non-degradable protein in MGM it is a good source of by-pass protein in ruminant diets (Chalupa et al., 1999). Schwab and colleagues (1976) found that methionine and lysine are the two main limiting amino acids for the production of milk. MGM mixed with either soya meal, or another lysine source, gives greater potential for MGM use in a dairy ration formulation.


Many oilseed crops produce a by-product meal or cake, which generally is a good quality protein source for livestock rations. Several common and some less common species that could be used in an expanded role as animal feeds are presented in Table 1. Several species of oilseeds, most notably soybeans and oilseed rape (canola) have been genetically modified to provide more specialised cultivars.

The increased demand for white meat around the world over the last 30 years has helped to fuel a large increase in the demand for high quality feedstuffs for these livestock sectors (Weiss, 2000). Intensive pig and poultry units are particularly sensitive to costs of production and have a need for high quality feeds, such as oilseed meals, to help keep feed conversion ratios low. The price of oilseed meal is related to the price the processor is able to get for the oil. The demand for a specific oilseed meal, therefore, is directly related to the oil price. Soya is one of the few oilseed crops that is an exception to this principle because of the low oil yield. Another factor that affects the value of a specific oilseed meal is its protein content.

Weiss (2000) outlines the relationship between oilseed meals and intensive livestock production. This will continue to encourage increased production of oilseeds meals as a protein source.

As is clear from Table 1, soya meal is the dominant meal source of protein for livestock. In 2000, Argentina and Brazil accounted for 61 percent of world exports of soya meal followed by the United States at 16 percent (USDA, 2002). The United States produces 34.8 Mt soya meal, but consumes 28.5 Mt. China produces 13.3 Mt but consumes an additional 11.5 Mt of soya meal. This additional soya meal is needed to supply China’s burgeoning livestock industry (USDA, 2002).

World production (Mt) of selected oilseed meals (million tonnes)

























Palm kernel





Canola (Rape)


















1 211.1

17 360.97






Source: (Weiss 2000; USDA, 2002)

Crambe (Crambe abyssinica), also known as Abyssinian kale, is a close relative to mustard and rape. It prefers sandy loam soils, but will also grow on a wide range of soils. Crambe grows well with similar cultivations to most small grains. Crambe meal cake typically contains 400-600 g/kg crude protein (fat free dehulled cake being 500-550 g/kg protein), with a good amino acid balance and a high lysine content (Weiss, 2000). Crambe can provide 1200 kg of dehulled meal per 1000 kg of oil.

Canola, oilseed rape (Brassica napus and Brassica rapa) meal provides a good protein source (430-450 g/kg crude protein [CP] - dehulled) with an excellent balance of essential amino acids. Some variation in the protein content of canola can be due to cultivar, soil type, and environmental factors (Bell, 1995). The lysine content of canola tends to be lower then soya meal, but canola has a higher proportion of sulphur amino acids

The major drawback to members of the brassica family is the presence of glucosinolates (hydrolysed to thioglucosidease) within the seed. These compounds can be toxic especially to non-ruminants, and will also reduce the palatability of the meal. The thioglucosidase and sinapine have to be extracted with hexane to allow crambe to be safely used for feeding cattle and poultry. In the case of canola, breeding work has been done to look for varieties with lower glucosinolates. These lower glucosinolates varieties tend also to have higher sulphur amino acids contents than other varieties (Bell, 1995). Based on current breeding trends, Bell (1995) suggests that future canola varieties will have very little glucosinolates.

Weiss (2000) describes crops like jojoba (Simmondsia chinensis) and niger (Guizotia abyssinica) as being able to provide an adequate protein supply for livestock feed. Jojoba, a crop found in north western Mexico and southern California, provides an extracted meal with 250-350 g/kg CP. However it needs to be detoxified through a one-stage oil extraction and meal detoxification process. Weiss (2000) goes on to say that the lysine content of jojoba is adequate but the methionine content is very low.

Niger grown in Ethiopia and India provides an extracted cake that can contain an average of 340 g/kg CP, without any anti-nutritional agents. Based on very few samples there tends to be low concentration of lysine and threonine in niger cake. In India, niger meal is used extensively as a feed for lactating cattle and buffalo (Weiss, 2000).

Safflower (Carthamus tinctorius) meal is not very suitable for monogastric animals, but can be fed to cattle and sheep. Safflower meal has 200-240 g/kg crude protein (Mündel et al., 2000). Weiss (2000) says that higher growth rates are possible from safflower than from a similar amount of soya meal in a properly balanced ration. Safflower meal can make a suitable supplement for the fattening of cattle. If safflower has been hulled it can be added to pig rations, replacing up to 25 percent of the ration’s protein requirements.

Sesame (Sesame indicum) meal provides an adequate protein source for livestock, (351-470 g/kg CP), but it must be used rapidly to prevent it from becoming rancid and unpalatable (Weiss, 2000). It is also valued as a source of protein in human diets, and so is less frequently used in animal rations. Sesame has a similar protein content to cottonseed meal and is also high in calcium and phosphorus. It is low in lysine, so must be fortified with soya meal if fed to pigs or poultry. There can be a reduction in the availability of calcium, magnesium, and zinc, because sesame contains 50 g/kg phytic acid (Weiss, 2000). There is a possibility that if too much sesame is included in the ration it could result in soft butter.

Cotton (Gossypium spp.) meal is palatable and commonly used in cattle rations in cotton growing regions of the United States. Solvent extracted cottonseed meal is the most common type of meal and has about 440 g/kg protein, similar to soya meal (, 2001). Cottonseed meal contains gossypol, a polyphenolic aldehyde, which can make cottonseed meal toxic to monogastric animals (McDonald et al., 1995). In ruminants it is unlikely that enough cottonseed meal would be ingested to result in the animal suffering from gossypol toxicity. Cottonseed meal is low in rumen degradable protein, and as such it is a suitable source of bypass protein (FAO, 2001a).

Sunflower (Helianthus annuus) meal has a high protein content, but because of the high fibre content of the seed husks, has to be de-hulled to achieve a 400 g/kg crude protein content (Weiss, 2000). As such it is of limited value to feed compounders. Sunflower meal is low in lysine and must be fortified by soya meal if it is to be used for feeding to pigs and poultry. When sunflower meal is used for feeding ruminants and horses it is generally mixed with grain and roughage. It can be fed in large amounts, up to 5 kg, to milking cattle without an adverse effect on the milk yield or quality (Weiss, 2000).

It is also possible to ensile whole plant sunflower, which will provide silage of about 140 g/kg crude protein. Sunflower grown for silage should be cut when half the heads are in bloom (FAO, 2001b). Sunflower for silage is often grown in areas that are too cool for growing maize (Zea mays).

Linseed (Linum usitatissimum) meal is the by-product of extracting the seed for oil. The meal contains 350-380 g/kg CP that is low in protein quality, being deficient in lysine. It has been a favourite protein source for horses and ruminants in the past. Today, soya meal is preferred as it is cheaper and of higher protein quality. The meal fed in large amounts is laxative, and excess amounts in rations have undesirable softening effects on butterfat and give milk a rancid taste. The recommended maximum intake for cattle is 3 kg per day. Because of this softening property of the oil, linseed cake is unsatisfactory as a main ingredient in pig feeds. Up to 1 kg per day has been used with good results, but not more than 8 percent linseed meal is commonly included in rations. For young pigs and brood sows a maximum of 5 percent linseed meal in the ration is usually recommended. Linseed meal is toxic to poultry except in very small proportions (under 3 percent). Larger amounts depress growth. The toxicity can largely be eliminated by soaking the meal in water for twenty-four hours or by adding pyridoxin, one of the B-vitamins to the diet (FAO, 2001c).


Groundnut (Arachis hypogaea) meal is a valuable feed that can provide 450-550 g/kg crude protein. The quality of the meal depends upon extraction method, cultivar, and where the crop was grown. Groundnut protein can be of comparable quality to soya meal. If groundnut meal is to be used for poultry diets, lysine and methionine have to be added to the ration. If too high a rate is included in pig and cattle rations it can result in soft pork and reduced milk fat quality (Weiss, 2000). However it can be fed to fattening cattle without any effect on meat quality.

Soybean (Glycine max) is by far the most dominant protein crop for livestock rations throughout the world. Generally soya meal can provide 440 to 480 g/kg crude protein. Soybeans provide a high quality and highly digestible protein source that is also high in lysine, making it well suited to feed compounders. Generally inclusion rates of soya meal in monogastric diets range from 30-40 percent. Because of the presence of trypsin inhibitor in the bean, it has become standard practice for the beans to be heated for an extended period of time before and during oil extraction.

Whole roasted soybeans can also be included in ruminant rations. While lower in protein that soy meal at 370 g/kg, roasted soybeans are higher in rumen bypass protein. Studies by the United States Department of Agriculture (USDA) suggest that higher milk yields are possible when roasted soybeans are substituted for soya meal in the diets of lactating dairy cattle (Satter and Dhiman, 1996).

It is also possible to use soybeans for forage production. When soybeans were initially introduced to North America they were grown exclusively as a forage crop. In recent years several new cultivars of forage type soybean have been developed (Devine and Hatley, 1998). In the United States, these varieties have achieved yields of 11 t DM/ha at 170 g/kg CP (Borman, 1998). Nutritionally, whole crop soybeans provide similar levels of protein and digestibility as alfalfa (Hintz et al., 1992).

The name lupin applies to a collection of four agriculturally important species. The two most important being Lupinus albus and Lupinus angustifolius offering relatively high yields (320-400 g/kg) of crude protein. Lupins tend to be deficient in sulphur containing amino acids and are also low in lysine (Haq, 1993). This means that in pig rations, lupins would need to be either supplemented with specific amino acids or with other protein crops. For pigs, lupins (ground seed) can have inclusion rates as high as 20 percent. For sheep and cattle they can be the sole concentrate protein feed. Sweet varieties of lupins should be used to avoid any problems with alkaloids. In Australia, lupins (L. angustifolius) are increasingly being used as a replacement for fishmeal and soya meal. Yields of lupins can vary greatly depending on environmental factors and to a lesser degree on species, with L. albus yielding up to 5 t DM/ha in France and Chile, and L. angustifolius up to 4 t DM/ha in New Zealand.

Lupins can be used as fresh forage or ensiled with maize or other cereals. All species have a similar chemical composition.

In temperate climates, peas (Pisum sativum) and beans (Vicia faba) provide a good source of home grown protein, with peas containing 250 g/kg crude protein and beans varying between 260 and 300 g/kg. The lysine content of peas and beans is lower than in soya meal. Their relatively high protein content and level of lysine mean that they are complementary to cereals. The protein is low in sulphur amino acids and tryptophan. Consistent yield can be a problem because of plant stress during flowering and pod filling. Protein content can also vary due to environmental conditions.

The European Commission (1994) estimate that if the rest of the European Union compounders used peas to the same extent as in France, the total use of peas would be 11.3 million tonnes, which is double current production levels. However, the best agronomic conditions to grow peas are found in France where they develop rapidly. The cost of production in France is now nearly equivalent to that of cereal crops. Peas and beans contain anti-nutritive factors - tannins and trypsin inhibitors. Pea varieties grown for animal feed are tannin-free and are screened for trypsin inhibitor activity. White flowered, tannin-free varieties of beans are also available. New varieties are screened for vicine and convicine, which cause haemolytic anaemia. These anti-nutritive factors are relatively unimportant for ruminants, and heat treatment destroys them during processing. (Shaw et al., 1998).

Forage peas can be a valuable source of protein for ruminant diets. Peas have higher basic CP and a lower non dietary fibre (NDF) value than ryegrass, suggesting they should have a higher dietary intake than ryegrass. Peas also have a lower metabolizable energy (ME) value, as expected, than ryegrass. Digestibility and overall feed value will be reduced if harvest is delayed beyond the mange tout stage (growth stage 205-206). Conversely, earlier harvesting can give a high-quality forage but at the expense of production. Arable silage will generally have lower crude protein (100-140 g/kg) DOMD (58-63 percent) and ME (9.0-10.0), although these values can be higher depending on the pea cultivar chosen and its contribution. In addition, the partner cereal species and cultivar, with respect to its straw length in particular, (with a long length being advantageous) will have an effect on digestible organic matter digestibility (DOMD) and ME.



In recent times the higher feeding value of legumes relative to grasses has been increasingly exploited, as a result of their higher nutritive value and increased animal intake. The rapid particle breakdown in the rumen resulting in rapid passage through the animal’s digestive system helps to contribute to the higher intakes (Waghorn et al., 1989). As a result of this high rumen degradation, those species that do not use condensed tannins to bind some of the protein can result in the fed animal suffering from bloat. This is only a serious problem in systems where animals are allowed to graze pastures that have high levels of low tannin forages. McMahon and colleagues (1999) found that including up to 20 percent sainfoin (DM basis) in fresh forage fed to cattle reduced the incidence of bloat by up to 93 percent, without negatively effecting nutritional quality of the feed.

Since the 1950s, the recognition of nitrogen as a key factor in grassland production has led to increased applications of fertilizer. The preference for nitrogenous fertilizers made other forms of nitrogen acquisition, including legumes, less popular and significant. (Frame et al., 1998). Eventually, with increases in the cost of inorganic nitrogen, greater interest has been shown in the potential of forage and grain legumes.

Forage legume prospects

The following examples represent species considered as forage legume prospects, with a potential role for further increasing future protein supply for livestock.

Bird’s-foot trefoil (Lotus corniculatus) is native to Europe, North Africa and parts of Asia. It is widely used for pasture, hay and silage production in areas where the soils or climate are not suitable for lucerne production. Pastures of bird’s foot trefoil should be allowed to reseed themselves in the late summer to ensure longer stand life (Beuselinck and Grant, 1995; Blumenthal and McGraw, 1999). Well maintained stands of bird’s-foot trefoil can be viable for over twenty years.

There are two distinct types of bird’s-foot trefoil, Empire and European. Cultivars of the Empire type are derived from naturalised ecotypes found in New York. European cultivars come from ecotypes found throughout Europe. The Empire type tends to have finer stems, to be prostrate, later flowering, indeterminate, more winter hardy and with slower seedling growth than the European type (Beuselinck and Grant, 1995).

The use of bird’s-foot trefoil in many temperate systems is limited by the crop’s very slow establishment, which is in part a result of the smaller seed size compared with lucerne and red clover (Beuselinck and Grant, 1995). It competes poorly with weeds or some companion crops (Frame et al., 1998). A great deal of research has recently improved our understanding of how to establish this crop. Once bird’s-foot trefoil is established it provides an excellent source of protein for livestock, with a nutritional value equal to if not greater than lucerne (Marten and Jordan, 1979). The crude protein content of the forage is dependent on the stage of development, but decreases at a slower rate than other forage legumes. At the early bloom stage, trefoil can provide around 210 g/kg CP. This crop also contains condensed tannins that help to prevent protein being metabolized in the rumen. In a review by Beuselinck and Grant (1995) they report that the cell wall of bird’s-foot trefoil also breaks down more slowly in the rumen than clovers and lucerne, increasing the amount of protein available in the lower intestine. These two factors that increase the bypass protein value of the crop help to make it a higher quality forage than some other legumes.

Lucerne (Medicago sativa) is often referred to as the queen of forages because of its ability to provide consistently high yields of high quality. Lucerne is the most widely grown forage legume in the world with 30 M ha, 85 percent of which is in the United States, Commonwealth of Independent States, Argentina, Canada, China, and Italy (Frame et al., 1998). It produces the greatest yield of protein per hectare of any of the temperate crops, including grains and oilseeds (Barnes and Sheaffer, 1995). It is most frequently used in conjunction with forage maize in dairy systems, because the protein of the lucerne complements the high energy maize. It is rarely grazed as a pasture crop because of the risk of bloat and possible death. It is generally harvested for hay or silage. The difficulty of producing good quality silage from lucerne has limited its use in some areas of the world, specifically maritime climates. Protein values for lucerne are also dependent upon the growth stage at which it is harvested and are generally around 200 g/kg, but have been reported to range from 129 to 324 g/kg (Spedding and Deikmahns, 1972).

Red clover (Trifolium pratense) is grown on over 4.5 M ha in the United States, making it the second most important forage legume grown in that country (Taylor and Smith, 1995). It is also of great importance in European farming, making it one of the most widely spread species in temperate agriculture. Red clover has adapted to a wide range of soil and climatic conditions, and tolerates growing on soils too acidic for lucerne production. It is however more prone to disease problems in climates that have higher summer temperatures. Red clover grows well in environments with sufficient moisture throughout the growing season (Taylor and Smith, 1995).

Red clover tends to be a very short-lived forage legume, perhaps as short as two years. However this tendency to be short-lived has been of increasing importance in some regions because of a move away from the use of mineral nitrogen, in both organic and conventional systems. It is able to fix high rates of nitrogen in a relatively short period, providing the opportunity to grow subsequent crops with little mineral nitrogen.

As a forage, red clover is very suitable for many ruminant systems, providing yields of up to 10 t DM/ha/year (Wilkins et al., 2001). Wilkins and colleagues (2001) managed to produce a red clover silage with 190 g/kg crude protein. In a review, Taylor and Smith (1995) report that the crude protein of red clover dropped from 280 g/kg in vegetative forage to 140 g/kg in full-bloom forage. Red clover also has greater digestibility than lucerne or bird’s-foot trefoil.

Sainfoin (Onobrychis viciifolia) was widely grown in Europe during the 17th to 19th centuries and to a lesser extent in the early 20th century. It was used as a source of very high quality hay, much of which was fed to heavy working horses of the time. The aftermath grazing was highly favoured for fattening lambs. There are two main types of sainfoin - the ‘common’ and ‘giant’ type. The common type lasts longest, whereas giant sainfoin is productive over a much shorter time span, but is more popular with the equine industry in the eastern counties of the United Kingdom. Sainfoin prefers calcareous soils with a pH of over 6. Several reports suggest that it is more drought tolerant than lucerne, and better suited to shallow brashy soils.

Sainfoin has many positive characteristics as a forage. In ruminants, these are the result of its high content of condensed tannins. These bond to the protein in the rumen and allow it to pass into the abomasum where it is digested. Daily live weight gains for cattle and lambs are high on grazed sainfoin. Agronomically, its positive characteristics include a deep taproot that allows the plant to be resistant to drought and of course, being a legume there is a high level of residual fertility after a sainfoin ley has been ploughed in. However, sainfoin does not persist well. This coupled with its rather low dry matter yields, the decline in the use of horses for farm work and the availability of cheap nitrogenous fertilisers, brought about a decline in the growing of this crop. Increasing interest in horse ownership for recreational purposes, however, could renew interest in the United Kingdom. Sainfoin is less efficient than lucerne and red clover at fixing gaseous nitrogen, and this may be one of the reasons for low persistence of sainfoin in many forage stands. Research has also shown that sainfoin invests less assimilates from photosynthesis in leaf production than lucerne. Frame and colleagues (1998) suggest that this constraint on energy availability might be one of the causes of the poorer nitrogen fixation in sainfoin.

With a move away from mineral nitrogen to what are considered (by some) to be more benign sources (forage legumes), sainfoin has an opportunity to make a comeback in suitable cropping systems. Several experiments have been performed to determine the suitability of sowing multiple legume mixtures that include sainfoin. These include mixtures with white clover, bird’s-foot trefoil and lucerne. The bird’s-foot trefoil mixture worked well when used in a conserved forage system. Sainfoin and lucerne have been sown together in the hope that the sainfoin could help reduce the risk of bloat from feeding cattle fresh lucerne. Experiments in Canada have shown it is possible to reduce the risk of bloat by feeding the two crops in a mixture (McMahon et al., 1999). Possible concerns with this mixture are that the sainfoin might not persist under the higher levels of competition from lucerne.

Yields of sainfoin have been found to be as high as 14 t DM/ha/year (Lane and Koivisto, 2000). This yield coupled with protein values ranging between 179 and 134 g/kg for full bloom forage, means that sainfoin can provide a good protein source in the right farming system. Meissner et al. (1993) also found that sainfoin delivers 50 percent more non-ammonium nitrogen to the small intestine than lucerne.

White clover (Trifolium repens) grows over a very wide range of agricultural lands including temperate and subtropical regions (Pederson, 1995). It accounts for 15 M ha of pastureland in Australia and 5 M ha in the United States. Its value to production systems in western Europe has been reappraised over the last 15 years, prompted by a desire to seek systems of production that are lower costing and more environmentally sustainable (Frame et al., 1998). White clover is generally grown in association with perennial ryegrass, but can also be grown with other forage grasses.

The yield potential of ryegrass/white clover mixtures has been estimated to be 18.5 to 22.5 t DM/ha/year in the United Kingdom and 22 to 28 t DM/ha/year in New Zealand (Frame et al., 1998). Yields more frequently tend to range between 7 and 15 t DM/ha/year in these countries. While it is generally grown as a pasture crop it is possible to make silage from some cultivars of white clover. Recent work by Wilkins et al. (2001) has produced silage with 250 g/kg of crude protein. This higher protein value for white clover tends, as a result of its growth habit with the main stem low to the ground, to be found in the leaf (Pederson, 1995) which has a higher proportion of protein than the stems. Cattle being fed this silage ad libitum were producing 2 kg/d more milk than those fed lucerne, red clover, or white clover mixed with ryegrass (Wilkins et al., 2001).

To protect the stolons of white clover, it is best if it grazed to no less than 5 cm above the soil surface and then given approximately 4 weeks to recover before re-grazing (Pederson, 1995). There should also be no more than 20 to 40 percent white clover in the pasture to prevent bloat.

Kudzu (Pueraria lobata) native to Japan, Korea, and China, is a vining legume that can be used for pastures and hay production in warmer climates. It does not reportedly tolerate close grazing and, therefore, like most species will work best in rotational grazing systems. The vines can grow up to 20 m in a season (Miller and Hoveland, 1995). Kudzu hay usually has a crude protein content of 150 to 180 g/kg (Everest et al., 1999). Kudzu unfortunately only yields between 4.5 and 9 t DM/ha/year, greatly limiting its potential as a forage crop in intensive livestock production systems.

Sericea Lespedeza (Lespedeza cuneata) is a warm seasoned perennial forage legume that is native to China, Korea and Japan (Hoveland and Evers, 1995). With the advent of new cultivars of sericea it might find a place in beef production systems. Older cultivars of the species tend to have a tannin level making them unpalatable to livestock, thereby reducing intakes and growth rates. Grazing trials have shown that low tannin sericea is inferior to lucerne, but superior to high tannin sericea. The tannin content decreases as the forage is dried for hay, making it more palatable to cattle in this form. The reported crude protein content is between 110 and 160 g/kg. If not grazed intensively, sericea can also quickly become dominant in a sward, crowding out many grasses. Sericea’s place in production systems is as a forage legume where other species are not a viable option.

Tropical legumes - herbaceous and tree species

There is a wide range of tropical legumes available to agriculturists as potential feed sources. Specific reference will be made to some example species (Quesenberry and Wofford, 2001), including Aeschynomene, Arachis, Centrosema, Desmodium, Leucaena, Macroptilium and Stylosanthes.

Early work on tropical forage legumes tended to revolve around plot evaluations, where too much emphasis was probably given to the yield potential and less consideration given to the importance of stand persistence under grazing, and competition from aggressive tropical grasses (Kretschmer and Pitman, 2001). Nevertheless, the high yield potential of these species encouraged researchers to continue breeding programmes. A considerable amount of the current tropical forage legume breeding and agronomy progress has been stimulated by the value that some of these species have in sub-tropical regions, of either Australia or the United States.

Aeschynomene is native to the tropical regions of the Americas and generally grows in an erect habit to a height of between 1 and 2 m. Most of the species are annual and are self-regenerating. None are known to be toxic to cattle. A. americana has a relatively high yield of protein at 150 to 250 g/kg in the leaves (Kretschmer and Pitman, 2001).

Rhizoma peanut (Arachis glabrata) has a yield potential of up to 10 t DM/ha/year when established and of 140 to 180 g/kg crude protein forage. The crop is established by planting dormant rhizomes and this is the key limiting factor to its expanded use despite there being no reported disease or nematode pest in the literature (Quesenberry and Wofford, 2001). Once established the rhizoma peanut has been known to persist in excess of 40 years, making it a very good choice for tropical production systems.

The genus Centrosema has 32 named species, most of which are twining perennials with trifoliate leaves, with or without stolons (Kretschmer and Pitman, 1995). Of these - C. acutifolium is limited geographically to southern Columbia, Venezuela and north-western Brazil, in areas that get in excess of 1000 mm/year of rain. It can withstand acid soils (pH as low as 4.3) with high aluminium and manganese. C. acutifolium is used as forage in grazed pastures or in cut-and-carry systems. Optimal cutting intervals depend on soil moisture and fertility, and a 10-14 week cutting interval and a 10-15 cm cutting height are suggested (‘t Mannetje, 2001). It is an effective under-storey for timber trees and in coconut plantations with about 60 percent light transmission. A cultivar 'Vichada' has been released in Colombia (‘t Mannetje, 2001). ‘t Mannetje (2001) goes on to say that on high fertility soils C. acutifolium will yield as high as 5 t DM/ha every 12 weeks; but this falls to 3 t DM/ha on poor wetter soils. C. acutifolium was more productive and persistent than the other Centrosema spp., regardless of companion grass, and this was associated with higher live weight gains particularly during the dry season (Lascano et al., 1989). In association with gamba grass, a live weight gain (LWG) of 180 kg/steer/year has been measured, as compared with 110 kg/steer/year from gamba grass alone. Milk yield in association with gamba grass was increased by 15 percent (1.2 kg/day) compared to grass alone (Lascano and Avila, 1991). For cut-and-carry systems, however, it is inferior to C. macrocarpum because of lower DM production (‘t Mannetje, 2001). The crude protein is similar to C. puescens at 624 g/kg.

C. macrocarpum is a more disease tolerant and diversely spread member of the genera (Kretschmer and Pitman, 1995). C. macrocarpum is used as forage, as ground cover in plantation agriculture and as a green manure. Similar to C. acutifolium it can either be grazed or cut. It also has the same nitrogen concentration in its leaves as C. acutifolium.

C. pubescens is a highly palatable forage legume that is the most commercially widespread member of this genera, and is used in both grazing systems and as a cover crop in plantation crops. Like the other two Centrosema spp., it can have 624 g/kg of nitrogen in the leaves (FAO, 2001d). It performs well in mixed swards with grasses, but cannot tolerate low pH and high soil aluminium or manganese (Kretschmer and Pitman, 1995). A yield potential of up 15 t DM/ha/year was reported in Columbia.

There are about 300 Desmodium species, with several key species being used successfully as forage crops. D. heterocarpon is native to south east Asia, Australia and the Pacific islands. It is a long-lived perennial, with a creeping stem. It is a highly persistent legume and grows well as a companion with tropical grasses such as Bahia (Paspalum notatum). Although it is lower in quality than some other tropical legumes, its persistence makes it a better choice. The crop performs best with a soil pH of between 5.0 and 6.0.

Macroptilium atropurpureum is a deep-rooting perennial with trailing pubescent stems which may root anywhere along their length, especially in moist clays but rarely in drier sandy soils. It contains 168 g/kg crude protein and should be lightly grazed at all times. Livestock will eat the runners back towards the crown, which should be protected from overgrazing. The concept that ‘leaf begets leaf’ is valid for M. atropurpureum, and grazing to 15 cm maintains the stand. In thinning stands, M. atropurpureum should be allowed to seed and for seed to shatter, so that new seedlings can improve the population density. In this way it will also climb over dominant grass and weeds, and suppress them. Stobbs (1969) found that a rotational grazing system of two weeks grazing-four weeks rest maintained the best botanical composition, and could equal the weight gain obtained for continuously grazed animals.

Stylosanthes spp. have been used very successfully for forage production throughout tropical and sub-tropical regions. They grow as either herbaceous or small shrub perennials that do not tolerate water logging. S. guanensis is used either for grazing or as a cover crop for plantation crops (Kretschmer and Pitman, 1995). Greatest yield potential for S. guanensis is in a rotational system, where it is grazed for one week and then given four weeks rest. It has a whole plant crude protein concentration of up to 181 g/kg. All of the Stylosanthes spp. tend to suffer badly from disease pressures, which limits their potential in a broader commercial context. Breeding work is needed, in particular, to improve disease resistance (Quesenberry and Wofford, 2001) and to improve tolerance to water logging.

Leucaena leucocephala is a native of Central America, growing as a shrubby tree to about 8 m in height, and yielding 116 g/kg CP. It is frequently grown in rows about 1.5 m wide with grass in between (Kretschmer and Pitman, 1995). Because of the crop’s ability to grow in a wide range of soil and climatic conditions, it is one of the most widely commercialized tropical legumes. It tends to develop slowly and is generally given a year to establish before grazing. Leucaena contains a toxic amino acid (mimosine) that means it should not be used in non-ruminant diets, but is suitable for tropical ruminants that have a rumen microflora capable of detoxification. This microbe has been isolated and transferred to ruminants in Australia (Kretschmer and Pitman, 1995), expanding the benefits of Leucaena to more farming systems. The leaf yield of Leucaena has been recorded as high as 26.8 t DM/ha in Fiji (Blair et al., 1990).

Blair and colleagues (1990) consider tree legumes to have the greatest potential to improve the supply and quality of protein in the human diet of the developing world, by providing a major source of fodder for livestock. There are recommendations that tree legumes should be used at inclusion rates of 30-50 percent in ruminant rations (Devendra, 1992), but because of anti-nutritive factors they should perhaps be used to a lesser extent in monogastric diets (D’Mello, 1992). Some of these species have condensed tannins (procyanidins and proanthocyanidins) that at lower concentrations can improve protein-use efficiency in ruminants. However, Calliandra calothyrsus contains 110 g/kg of condensed tannins, and at this concentration it can greatly reduce protein digestibility. Up to 36 percent inclusion in forage could be considered (Ahn et al., 1989).

Humphreys (1994) considers that many farmers in the tropics prefer to use tree legumes for cutting and carting to livestock or for direct grazing, rather than incorporating into the soil as a green manure to boost grain yields. Their preference is based on a belief that there are higher economic returns from the animals, than from the grain grown with green manures. Humphreys (1994) suggests that a more effective activity would be to supplement the legume fodder with either maize or maize husks, and then use the animal manure to further support grain crop production.

Quality Protein Maize

In 1963 scientists in the United States discovered the gene called opaque-2, which improves the nutritional quality of maize by increasing its lysine and tryptophan content. Initially farmers showed little interest in opaque-2 maize because of its low yields, chalky-looking grain, and susceptibility to pests and diseases. But in 1970, with funding from the United Nations Development Programme (UNDP) the arduous task of breeding maize to overcome these drawbacks was initiated.

During the 1970s and 1980s UNDP provided US$17 million in funding to the International Maize and Wheat Research Center (CIMMYT) in Mexico for this research. CIMMYT breeders continued to build on these efforts in the 1990s with support from the Nippon Foundation. CIMMYT has worked closely with research services in developing nations to breed varieties of quality protein maize (QPM) that are well adapted to local environmental conditions. Derivatives are now grown on approximately 1 M ha in 22 developing nations, and there are estimates that this will expand to 3.5 M ha by 2003.

QPM has come a long way since 1963. It now looks, tastes and yields like normal maize, but has nearly twice the levels of lysine and tryptophan - essential dietary amino acids - in its grain, as normal maize (Córdova, 2001). In normal maize the prolamine-type amino acids (which are not digestible) predominate. In QPM, the protein comprises 40 percent highly digestible glutelins and a balanced leucine-isoleucine ratio that boosts the production of niacin when eaten. QPM has better food and feed efficiency ratings (e.g. food intake/g weight gain) than normal maize (Córdova, 2001). QPM can have as much as 135 g/kg of crude protein and 35 g/kg more protein than normal cultivars of maize (Córdova, 2001).

A synopsis of alternative plant protein sources is given at Appendix A

Tropical Americas

With an ever-expanding agricultural frontier in tropical America, livestock are being displaced toward areas with low-fertility soils (mainly oxisols and ultisols). Meat and milk production in these areas is limited by the poor quality of available grasses and legumes. Lascano (2001) says that adoption of legumes for livestock production in tropical America has been poor, and many are not aware of the benefit these species can provide. He argues that the best way for this to be improved is through increased emphasis of ‘on-farm’ participatory evaluations of grass-legume pastures, and the demonstration of commercial benefits. National research programmes have recently, with collaboration from the Tropical Forages Project at CIAT, released new forage cultivars and appropriate technologies for their establishment and successful management.

The tropical Americas are the source of origin for many of the tropical legume species, but work has been restricted to relatively few. This restriction of potential plant diversity work has limited the utilization of these species in livestock production areas in tropical America (Lascano, 2001). It has led to farmers not seeing these prospects as being a potentially valuable part of their production systems. In Columbia this has recently been redressed by the introduction of a new high quality cultivar of perennial forage peanut, A. pintoi, (cv. Mani Forrajero). This cultivar is able to compete well with aggressive grasses (Gorf, 1985) and heavy grazing (Lascano, 1994).

As in temperate pastures, tropical legumes offer higher nutritive value compared with grasses, and can also provide a residual nitrogen source for the companion grass (Lascano, 2001). The nitrogen economy within pastures is highly dependent upon the legume content of the pasture. This premise is of equal importance in both tropical and temperate agriculture. CIAT (1989) illustrated a dependence of LWG on the legume content of a pasture, with cattle gaining 0.7 kg/hd/d;this was confirmed by Lascano (1994) who was also able to double the weight gain of cattle grazing pastures that were mixtures of legumes and grasses rather than grass pastures alone.

Substituting feed concentrates for forages and nitrogen fertilization of grassland has driven much of the research into improving tropical milk production. Both methods have been effective in increasing milk production per unit area and per cow (Lascano, 2001), but they are more expensive than using legume grass mixtures. Little work has been done to determine the potential of using legume/grass swards for increased milk production per unit area. Work in Australia suggests that greater yields are available from using nitrogen fertilizer on grass swards, because it is possible to have higher stocking rates. However, the production per animal grazing mixed legume-grass swards was almost twice that of the nitrogen fertilized sward (Lascano, 2001).

Over a period of four years, root-feeding beetles reduced the legume proportion within a pasture initially containing 60 percent legumes. This deterioration resulted in a decline in LWG to 0.25 kg/hd/d, not only because of a lack of legumes for forage but also because of a reduction in the nitrogen concentration of the grasses in the sward (CIAT, 1989).

Legumes can supply not only a good source of protein for livestock in tropical systems, but can also provide a cheap source of nitrogen to support grass production. These crops can also improve soil organic matter through decaying nodules and leaf litter. This will help to support longer term farming on these lands, meaning that less new land (often valuable tropical forest) needs to be brought into production (CIAT, 1989). Many of the legume species also have a deep taproot enabling them to better deal with drought conditions than grasses in the dry season. Even with all these advantages, legumes still do not yet play a major part in tropical American agriculture. More work has be done to encourage farmers to use these species, and to incorporate them into their production systems. Lascano (2001) suggests this can be achieved through on-farm pasture evaluation and demonstration programmes, so that the farmer can see first-hand the benefits of using these species. He goes on to suggest that there should also be more training for pasture establishment and grazing management, and the development of a reliable seed supply system.

United Kingdom and European Union

General European Union concern over the environmental implications of grassland agriculture has led to a policy that seeks to decrease stocking rates for livestock, and to reduce the amount of mineral nitrogen that is used on grassland. These policies along with increases in the genetic merit of livestock, specifically dairy cattle, have been encouraging a shift towards the use of forage legumes as high quality feed, and to reduce the need for mineral nitrogen (Frame et al., 1998). The United Kingdoms’ agricultural economy is dominated by grassland agriculture and is therefore a good example to represent the northerly part of the European Union.

In the United Kingdom, the protein source has come into clearer focus as a result of restrictions on using animal proteins, efforts to reduce the costs of ruminant production and public concern arising from potential nitrogen pollution (Wilkins and Jones, 2000). The British response to BSE, specifically the resultant ban on animal proteins is well documented. This, together with the fact that only 5-20 percent of nitrogen consumed by ruminants was being recovered in meat or milk (Wilkins and Jones, 2000), has provided an impetus to search for alternative sources of protein for livestock production.

Wilkins and Jones (2000) contend that it would certainly be possible to increase, relatively cheaply, the protein supply from plant-based proteins because some of these sources are very inefficient in providing nitrogen conversion for animals. Doubling the nitrogen fertilization rates from 120 to 300 kg/ha on grassland could produce twice as much grass-based crude protein (Morrison et al., 1980). Because of the inadequacy of the protein supplied, however, the animals would not be able to ingest enough of this grass to obtain the nutrients they need for required production. Grass proteins are highly rumen soluble, and because of low energy supply, the microbial protein synthesis is limited. It will be necessary, therefore, to look for better ways of balancing protein and energy fractions in the rumen, and to increase the proportion of rumen undegradable protein (UDP) in the diet.

Including temperate legume forages in ruminant rations can help to provide these changes. All have high feed intake values and result in higher milk production levels (Dewhurst et al., 2000).

There are, however, clear differences between several of these species in protein composition. Winters and colleagues (1999) found that white clover and lucerne protein are highly soluble and can have a higher proportional breakdown into free amino acids when ensiled. Red clover is more resistant to proteolysis, meaning that microbial nitrogen synthesis is more efficient (at 34 percent [Davies et al., 1999]) than for lucerne. The resistance of red clover to proteolysis is attributed to an elevated concentration of polyphenol oxidase (Albrecht and Muck, 1991) also suggesting that tannins in sainfoin and bird’s-foot trefoil offer similar protein protection during ensiling and in the rumen. This rumen protection has helped to make these forages an excellent source of UDP. Several authors (Waghorn and Shelton, 1997; Thomson et al., 1971) have shown the benefits of sainfoin and bird’s-foot trefoil to protein utilization. Consequently this has resulted in higher growth rates from these species relative to other legumes (Hart and Sahlu, 1993; Karnezos et al., 1994; Marten et al., 1987; Ulyatt, 1981).

However, these species can be unreliable agronomically, tending to be slow to establish and having poor stand persistence. The nutritional attractiveness of the tanniferous forages should encourage further research into improving their agronomic properties.

There is also room to considerably expand the role of concentrated forms of protein in British agriculture, and the most likely sources for this are beans, peas and lupins. Canola (rapeseed) meal is already well established as a feed for pigs and other livestock in the United Kingdom. Peas and beans are well known to many farmers, but they are not thought by some to be a good enough concentrate alongside grass-based rations due to the high level of grass protein loss with rumen degradation. They might work better, however, in maize-based systems (Wilkins and Jones, 2001).

Lupins have almost the same proportion of their protein as UDP as soybeans. There is currently, unfortunately, very little work to compare lupins with soybeans and fishmeal in dairy rations, but interim results suggest that lupins could be a suitable replacement for some of these other sources (Mansbridge and Blake, 1998). There is also much more work needed on the agronomy in Britain before it can become a major crop.

Economic performance

In practice, the unit cost of protein is clearly important for the competitiveness and commercial viability of a livestock enterprise. For driving down production costs it is important to know the cheapest and most effective protein source.

Farmers, however, do not normally think about the unit cost of producing home-grown protein when planning cropping programmes and their farm businesses. They are usually more concerned about their whole farming system planning; farm resources; how well a crop fits into the rotation and in particular for arable legume crops, what the financial output reflected in the crop gross margin might be. In the same way, and from a similar broad farm business perspective, a livestock farmer may judge a home-grown crop mostly on the basis of the contribution it will/can make to the overall animal feeding programme. Protein cropping on a farm is not usually viewed from the perspective of only protein production per hectare.

The strengths and weaknesses, therefore, of home-grown protein crop production and farm use are rarely compared directly or on the same basis as purchased external animal feeds.

From an economic standpoint, most arable farmers will be heavily influenced not only by the output level of a legume seed crop, but also the consistency of the gross margin and its expected contribution to arable performance.

In the United Kingdom, the recent greater popularity of oilseed rape in comparison with peas and beans is attributed to its more attractive gross margin, greater support and consistent yield performance - apparent in Table 2 (ENTEC, 1998).

Gross margins and variable costs of crude protein production costs of example United Kingdom arable crops


Gross margin with subsidy (£/ha)-1

Gross margin without subsidy (£/ha)-1

Crude protein produced (kg protein/ha)-1

Variable costs of crude protein (pence/kg protein)-1

Winter Beans





Winter Peas










Winter Oilseed Rape





Spring Linseed





Winter Wheat





Source: ENTEC, 1998. UK £ = US$1.53

Studies in the United Kingdom show, in general, that the cheapest protein (as crude protein) is obtained from forage protein crops (ENTEC, 1998). Some of the less common forage crops having the lowest growing costs. White clover, which is more commonly grown in mixed grass swards, has a unit cost of production comparable to perennial ryegrass when grazed or cut.

The cost of producing protein in the United Kingdom (which is favourable to grass growth) from perennial ryegrass or white clover in a mixed grass sward, is about 75 percent of the cost of beans, which are the cheapest arable protein crop (ENTEC, 1998).


Crude protein production and variable costs of crude protein production of example United Kingdom forage crops


Crude protein production kg protein/ha-1

Variable costs of crude protein pence/kg protein-1

Perennial Ryegrass

1 344


White Clover/grass

1 216



2 320


Red Clover

2 155


Forage Peas



Maize Silage



Of the United Kingdom arable protein crops, peas are the cheapest source of crude protein after beans. The crops that could make a very big difference to providing protein at least cost, lupins and soybeans, are at present rarely grown commercially.


In seeking new protein sources it would be helpful to know why certain potential protein crops, which are not widely grown currently, are not more attractive. To understand in greater detail, for example, why crop uptake has not been greater by farmers. This currently applies not only to more minor (less significant) sources but to such specific crops as peas (e.g. in the United Kingdom) or Leucaena trees (e.g. in Malaysia). The reasons could be mainly technical but learning about farmer perception and the overall suitability of ‘new’ prospects for particular farming systems would also be of interest. Such a study could extend to feed manufacturers, to gain a better understanding of their needs and their willingness to give greater consideration to currently less popular protein sources for feed incorporation.

An appropriate strategy might then be devised, focussing on increasing the production and use of these ‘new’ sources.

As part of this perspective, more information will be required on the influences of variety, crop husbandry and season on the protein quality of particular crop species.

In this respect new feed strategies for particular purposes, utilizing alternative protein sources in particular, should exploit nutrition models to a greater extent. These should also include, where appropriate, an economic modelling dimension.

From a medium term research and development perspective, more work is required, particularly and as always, on improving crop performance. More basic agronomy is also needed to include better exploitation of nitrogen fixation and the inoculation of suitable legumes.

Breeding and or genetic manipulation for better plant protein quality has to be a continuing goal with emphasis on the medium to longer term. Although this may be difficult technically with certain crops, it should be given greater emphasis.

Particular efforts may need to be supported in terms of research strategies to provide a better basis for farming systems to adjust to climate change. Small farmers in some developing countries could be seriously affected by such change.

Production of protein in practice has considerable future potential, with no shortage of possible supply routes. Realising the potential, however, from research to farm and feed manufacturer will take some considerable investment and co-ordinated efforts. We are fortunate to have so many prospects but a collective vision and action strategy will be essential to secure the desired outcomes. Co-operation between a key United Nations agency, the Food and Agriculture Organization, and a private sector trade association representing the feed manufacturers, the International Feed Industry Federation, is a very encouraging and welcome contribution.


Ahn, J. H., Robertson, B. M., Elliot, R., Gutteridge, R. C. & Ford, C.V. (1989). Quality assessment of tropical browse legumes: tannin content and protein degradation. Animal Feed Science and Technology, 27: 147-156.

Albrecht, K. A. & Muck, R. E. (1991). Crop quality and utilisation: proteolysis in ensiled forage legumes that vary in tannin concentration. Crop Science. 31: 464-469.

Avery, D. T. (1998). It’s time for the world to get realistic about meat. Outlook on Agriculture, 27, No. (1),: 35-40.

Baines, R. N. & Davies, W. P. (1998) Quality assurance in international food supply. In: Proceedings of the Third International Conference on Chain Management in the Agribusiness and Food Industry, p. 2133-223. Wageningen, Netherlands.

Barnes, D.K. & Sheaffer, C.C. (1995). Alfalfa. pg. 205-216. In R.F. Barnes, R.F.,D.A. Miller, D.A. & C.J. Nelson, C.J. (eds). Forages: an introduction to grassland agriculture, 5th ed. p. 205-216. Ames, USA, Iowa State University., Ames, USA.

Bell, J.M. (1995). Meal and by-product utilisation in animal nutrition. pg. 301-337. In D. Kimber, D. & D.I. McGregor, D.I. (eds). Brassica oilseeds: production and Uutilisation., p. 301-337. CABI, Wallingford, UK,. CABI.

Beuselinck, P.R & Grant, W.F.(1995). Bird’s-foot trefoil. pg. 237-248. In R.F. Barnes, R.F.,D.A. Miller, D.A. & C. J. Nelson, C. J. (eds). Forages: An Introduction to Grassland Agriculture, 5th ed. p. 237-248. Ames, USA, Iowa State University, Ames, USA.

Blair, G., Catchpoole, D. & Horne, P. 1990. Forage tree legumes: their management and contribution to the nitrogen economy of wet and humid tropical environments. Advances in Agronomy, 44:, 27-54.

Borman, K. (1998). Forage soybeans: A crop for the future. Missouri Soybean Farmer. Online (available at ( verified March 18, 2001.

Conway, G. (1998). A doubly green revolution. Biologist, 45:, 85-86.

CEDAR. (1995). Mixed Forage Diets for Dairy Cows., Report of project undertaken at CEDAR., Reading, UK, University of Reading, UK.

Chalupa, W., Boston, R., Sniffen, C. J. & Miner W. H. (1999). Formulating rations for dairy cattle on the basis of amino acids. In: Proceedings of the Volac Conference, March 18, 1999, Bugbroke, UK.

CIAT. (1989). Legumes: The Key to Productive Pastures. CIAT Report, p. 55-57. Cali, Columbia, 1989, 55-57.

Córdova, H. (2001). Quality protein maize: improved nutrition and livelihoods for the poor. pg. 27-31. In, S. Pandey, S. (ed). Maize Research Highlights 1999-2000. p. 27-31. CIMMYT Technical Bulletin. 6815. Mexico.

Crawshaw, R. (2001). Co-product Feeds: Animal Feeds from the Food and Drinks Industries. Nottingham, UK, Nottingham University Press, UK.

Curry, D. (2002). Farming and food: a sustainable future. Report on the Policy Commission of Farming and Food to the UK Government. Cabinet Office and DEFRA. London, UK.

Davies, W. P. & Turner, J.C. (2002). The Modern Food Chain: Profiting from Effective Integration. In: Proceedings of a Ministry of Agriculture Malaysia and British High Commission Seminar on the Food Chain, p. 1-37. Kuala Lumpur, Malaysia.

Davies, D.R., Winters, A.L., Leemans, D.K., Dhanoa, M.S. & Merry, R.J. (1999). The effect of inoculant treatment of alternative crop forages on silage quality and in vitro rumen function. pg. 131-132. In, T. Pauly, T. (ed.) Proceedings 12th International Silage Conference, p. 131-132. Uppsala, Sweden.

Delgado, G., Rosegrant, M., Steinfeld, H., Ehui, S. & Courbois C. (2001). Livestock to 2020: the next food revolution. Outlook on Agriculture 30, No. (1): 27-29.

Devendra, C. and & Thomas, D. (2002). Crop - animal systems in Asia: importance of livestock and characterisation of agro-ecological zones. Agricultural Systems, 71, 5-15.

Devendra, C. & Sevilla, C.C. (2002). Availability and use of feed resources in crop-animal systems in Asia. Agricultural Systems, 71:, 59-73.

Devendra, C. (1992). Nutritional potential of fodder trees and shrubs as protein sources in rumen nutrition. In A. Speedy, A. and & P. L. Pugliese, P. L. (eds.) Legume trees and other fodder trees as protein sources for livestock. FAO Animal Production and Health Paper, 102, 95-113.

Devine, T.E., & Hatley, E.O. (1998). Registration of 'Donegal' forage soybean. Crop Science, 38:, 1719-1720.

Dewhurst, R.J., Davies, D.W.R., Fisher, W.J., Bertilsson, J. and Wilkins, R.J. (2000). Intake and milk production responses to legume silages offered to Holstein-Friesian cows. Proceedings of the British Society of Animal Science.

D’Mello, J.P.F. (1992). Nutritional potentialities of fodder trees and fodder shrubs as protein sources in monogastric nutrition. In A. Speedy, A. & Pugliese, P. L. (eds.) Legume trees and other fodder trees as protein sources for livestock. FAO Animal Production and Health Paper, 102:, 115-127.

ENTEC. (1998). Home grown proteins for animal feeds. Technical Report. Entec UK Limited, Leamington Spa, UK, ENTEC UK Limited.

European Commission Agro-Industrial Research Division. (1994). Ways to improve the quality of domestically produced proteins for use in the animal feed industry. European Commission, Brussels.

Everest, J.W., Miller, J.H., Ball, D.M. & Patterson, M. (1999). Kudzu in Alabama: history, uses and control. Report ANR-65. Auburn, USA, Alabama Co-operative Extension System, Auburn, USA.

Food Standards Agency. (2001). Submission to the Policy Commission on Food and Farming in England. London, UK, Food Standards Agency, London, UK.

FAO. (2001a). Gossypium spp. Cotton. Animal Feed Resources Information System. FAO. Online (available at ( verified March 18, 2002.

FAO. (2001b). Helianthus annuus Sunflower. Animal Feed Resources Information System. FAO. (available Online at ( verified March 18, 2002.

FAO. (2001c). Linum usitatissimum Flax, linseed. Animal Feed Resources Information System. FAO. Online (available at ( verified March 18, 2002.

FAO. (2001d). Centrosema pubescens Benth. FAO. (available at verified March 20, 2002.

Frame, J., Charlton, J.F.L. & Laidlaw, A.S. (1998). Temperate forage legumes. Wallingford, UK, CAB International., Wallingford, UK

Gill, M. (1999). Meat production in developing countries. Proceedings of the Nutrition Society, 58:, 371-376.

Gorf, B. (1985). Forage attributes of the perennial groundtnut Arachis pintoi in a tropical savanna environment in Columbia. pg. 168. In Proceedings 15th International Grassland Congress, p. 168. Kyoto, Japan.

Haq, N. (1993). Lupins (Lupinus species). pg. 103-130. In J.T. Williams, J.T. (ed.) Underutilized crops: pulses and vegetables. p. 103-130. London, UK, Chapman & Hall., London, UK.

Hart, S.P. & Sahlu, T. (1993). Mohair production and body-weight gains of yearling angora goats grazing forages with different tannin levels. pg. 575-576. In M.J. Baker, M.J. (ed.) Proceedings of 17th International Grassland Congress. Palmerston North, New Zealand.

Heap, R. B. (1998). Animals and the human food chain. In J.C. Waterlow, D.G. Armstrong; L. Fowden & R. Riley, eds. Chapter 17 Feeding a world population of more than eight billion people. Chapter 17.(Eds) J.C. Waterlow, D.G. Armstrong; L. Fowden, R. Riley. Oxford, UK, Oxford University Press, Oxford, UK.

Hintz, R.W., Albrecht, K.A., & Oplinger, E.S. (1992). Yield and quality of soybean forage as effected by cultivar and management practices. Journal of the America Society of Agronomy, 84:, 795 - 798.

Humphreys, L.R. (1994). Tropical forages: their role in sustainable agriculture. Harlow, UK, Longman Scientific & Technical, Harlow, UK.

IFIF (2002) Roger Gilbert pers comm. (2001). Cottonseed meal. Online(available at ( verified March 18, 2002.

James, C. (2001.) Global Rreview of Ttransgenic Crops: 2000. International Service for the Acquisition of Agri-Biotech Applications. Ithaca, New York, USA, International Service for the Acquisition of Agri-Biotech Applications, Ithaca.

Karnezos, T.P., Matches, A.G. & Brown, C.P. (1994). Spring lamb production on alfalfa, sainfoin, and wheatgrass pastures. Agronomy Journal, 86:, 497-502.

Kretschmer, A.E. & Pitman, W.D. (1995). Tropical and subtropical forages. pg. 283-304. In R.F. Barnes, R.F.,D.A. Miller, D.A. & C.J. Nelson, C.J. (eds.) Forages: An Introduction to Grassland Agriculture, 5th ed. p. 283-304. Iowa State University, Ames, USA, Iowa State University.

Kretschmer, A.E. & Pitman, W.D. (2001). Germplasm resources of tropical forage legumes. pg. 41-57. In A. Sotomayor-Rios, A. & W.D. Pitman, W.D. (eds.) Tropical forage plants: development and uses. p. 41-57. CRC Press, Boca Raton, FL., USA, CRC Press.

Lane, G.P.F & Koivisto, J.M. 2000. A re-assessment of the potential of sainfoin (Onobrychis viciifolia Scop.) as a forage crop for the United Kingdom. p. 202-205. In Proceedings of American Forage and Grassland Council and the North American Alfalfa Improvement Conference, p. 202-205. Madison, USA.

Lascano C., Estrada J., Avila P. (1989). Animal production of pastures based on Centrosema spp. in the eastern plains of Colombia. pg. 1177-1178. Proceedings. of the 16th International Grassland Congress, p. 1177-1178. Nice, France.

Lascano, C.E. & Avila, P. (1991). Milk production potential on pastures composed of grass alone or of grass combined with legumes suited to acid soils. Pasturas Tropicales, 13:, 2-10.

Lascano, C.E. (1994). Nutritive value and animal performance of forage Arachis,. In P.C. Kerridge, P.C. & W. Hardy, W. (eds.) Biology and agronomy of forage Arachis. CIAT, Cali., Columbia, CIAT.

Lascano, C.E. (2001). Animal production in grass-legume pastures in the tropics. pg. 219-232. In A. Sotomayor-Rios, A. & W.D. Pitman, W.D. (eds.) Tropical Forage Plants: Development and Uses. p. 219-232. CRC Press, Boca Raton, FL, USA, CRC Press.

‘t Mannetje, L. (2001). Centrosema acutifolium Benth. FAO Online (available at ( verified March 20, 2002.

McDonald, P., Edwards, R.A., Greenhalgh, J.F.D. and Morgan, C.A. (1995). Animal Nutrition (5th ed). Longman, New York, USA, Longman.

McGraw, R.L. & Hoveland, C.S. (1995). Lespedezas. pg. 261-271. In R.F. Barnes, R.F., D.A. Miller, D.A. & C.J. Nelson, C.J. (eds.) Forages: an introduction to grassland agriculture, 5th ed. p. 261-271. Ames, USA, Iowa State University, Ames, USA.

McMahon, L.R., Majak, W., McAllister, T.A, Hall, J.W., Jones, G.A., Popp, J.D. & Cheng, K.-J. 1999. Effect of sainfoin on in vitro digestion of fresh alfalfa and bloat in steers. Canadian Journal of Animal Science, 79: 203-212.

Mansbridge, R.J. & Blake, J.S. (1998). The effect of feeding different protein sources on intake, milk yield, milk composition and live weight in high yielding Holstein cows. pg. 202. In Proceedings of the British Society of Animal Production. p. 202.

Marten, G.C. & Jordan, R.M. (1979). Substitution value of bird’s-foot trefoil for alfalfa-grass in pasture systems. Agronomy Journal, 71, 55-59.

Marten, G.C., Ehle, F.R., & Ristau, E.A. (1987). Performance and photosensitization of cattle related to forage quality of forage legumes. Crop Science, 27: 138-145.

Martin, A. M. (2001). The future of the world food system. Outlook on Agriculture, Vol 30, N° (1), 11-19.

Meissner, H.H., Smuts, M., Vanniekerk, W.A. & Achinpongboateng, O. (1993). Rumen ammonia concentrations and non-ammonia nitrogen passage to and apparent absorption from small intestine of sheep ingesting subtropical, temperate and tannin-containing forages. South African Journal of Animal Science, 23, 92-97.

Merry, R.J., Jones, R. & Theordorou, M.K. (2001). Alternative forages - back to the future. Biologist, 48:, 30-34.

Miller, D.A. & Hoveland, C.S. (1995). Other temperate Llegumes. pg. 273-281. In R.F. Barnes, R.F., D.A. Miller, D.A. & C.J. Nelson, C.J. (eds.) Forages: An Introduction to Grassland Agriculture, 5th ed. p. 273-281. Ames, USA, Iowa State University, Ames, USA.

Millward, D.J. (1999). Meat or wheat for the next millennium? Proceedings of the Nutrition Society, 58, 209-210.

Minson, D.J. (1997) Ruminants: the protein producers. Biologist, 44: 463-464.

Morrison, J., Jackson, M.V., and Sparrow, P.E. (1980). The response of perennial ryegrass to fertiliser nitrogen in relation to climate and soil. Technical Report 27, Grassland Research Institute. GRI, Hurley, UK, Grassland Research Institute.

Mündel, H., Morrison, R.J., Blackshaw, R.E. & Roth, B. (2000). Safflower production on the Canadian prairies. Online (available at ( verified March 18, 2002.

Pederson, G.A. (1995). White clover and other perennial clovers. pg. 227-236. In R.F. Barnes, R.F., D.A. Miller, D.A. and C.J. Nelson, C.J. (eds.) Forages: An Introduction to Grassland Agriculture, 5th ed. p. 227-236. Ames, USA, Iowa State University, Ames, USA.

Pinstrup-Andersen, P., Pandya-Lorch, R., & Rosegrant, M.W. (1999). World Ffood Prospects. Critical Issues for the Early Twenty-First Century. Food Policy Report. Washington, USA, International Food Policy Research Institute, Washington, USA.

Quesenberry, K.H. & Wofford, D.S. (2001). Tropical forage legume breeding. pg. 81-105. In A. Sotomayor-Rios, A. & W.D. Pitman, W.D. (eds.) Tropical forage plants: development and uses. p. 81-105. CRC Press, Boca Raton, USA, CRC Press.

Ramirez, P. A. (1987). New pastures for the Colombian Plain. Pasturas Tropicales, 9:, 48-49.

Robinson, D. & Singh, D.N. (2001). Alternative protein sources for laying hens. Publication No. 144. Australia, Rural Industries Research and Development Corporation, Australia.

Rosegrant, M.W., Leach, N. & Gerpacio, R.V. (1999). Meat or wheat for the next millennium? Alternative futures for world cereal and meat consumption. Proceedings of the Nutrition Society, 58:, 219-234.

Sanders, T.A.B. (1999). Meat or wheat for the next millennium? The nutritional adequacy of plant-based diets. Proceedings of the Nutrition Society, 58. 265-269.

Satter, L. & Dhiman, T. (1996). Formulating dairy diet - enhancing profitability and reducing environmental impact. In Proceedings of 1996 Informational Conference for the Dairy and Forage Industries. Madison, USA, US Dairy Forage Research Center, Madison, USA.

Schwab, C.G., Satter, L.D. & Clay, A.B. (1976). Response of lactating dairy cows to abomasal infusion of amino acids. Journal Dairy Science, 59, 1254-1270.

Sere, C. & Steinfeld, H. (1996). World livestock production systems: current status, issues and trends. FAO Animal Production and Health paper No. 127. Rome: FAO, Rome, Italy.

Shaw, K., Allen, D., Adams, N., Burt, W., Rickard, S. & Taylor, K. (1998). Home grown proteins for animal feeds. Milk Development Council, Cirencester, UK, Milk Development Council.

Spedding, C.R.W. & Diekmahns, E.C. (1972). Grasses and Legumes in British Agriculture. pg. 515. Bulletin No. 49, p. 515. Commonwealth Bureau of Pastures and Field Crops. Farnham Royal, UK, Commonwealth Bureau of Pastures and Field Crops.

Stobbs, T.H. (1969). The effect of grazing management upon pasture productivity in Uganda. III. Rotational and continuous grazing. Tropical Agriculture Station Journal, 46:, 293-301. Augustine, WI, USA.

Taylor, N.L. and Smith, R.R. (1995). Red clover. pg. 217-226. In R.F. Barnes, R.F., D.A. Miller, D.A. and C.J. Nelson, C.J. (eds.) Forages: An Introduction to Grassland Agriculture, 5th ed. p. 217-226. Iowa State University, Ames, USA, Iowa State University.

Teferedegne, B. 2000. New perspectives on the use of tropical plants to improve ruminant nutrition. Proceedings of the Nutrition Society, 59, 209-214.

Thomas, D., Zerbini, E., Rao, P.P. & Vaidyanathan, A. 2002. Increasing animal productivity on small mixed farms in South Asia: A system perspective. Agricultural Systems, 71:, 41-57.

Thomson, D.J., Beever, D.E., Harrison, D.G., Hill, I.W. & Osbourn, D.F. (1971). The digestion of dried sainfoin and dried lucerne by sheep. Proceedings of the Nutrition Society, 30:, 14A.

Ulyatt, M.J. (1981). The feeding value of herbage: can it be improved? New Zealand Journal of Agricultural Science, 15:, 200-205.

USDA. (2002). Oilseeds: World Markets and Trade. Circular FOP 03-02, USDA, Washington, DC, USDA.

Waghorn, G.C., Shelton, I.D. & Thomas, V.J. (1989). Particle breakdown and rumen digestion of fresh ryegrass (Lolium perenne L.) and lucerne (Medicago sativa L.) fed to cows during a restricted feeding period. British Journal of Nutrition, 61:, 409-423.

Waghorn, G.C. and Shelton, I.D. (1997). Effect of condensed tannins in Lotus corniculatus on the nutritive value of pasture for sheep. Journal of Agricultural Science, 128, 365-372.

Warman, K.D; Conway, J.S; Goulding, K.W.T; Hillam, J; Jarvis, S.C; & Donaghy, D. (1997). Nitrogen flows in mixed farming systems. In Proceedings of the Association of Applied Biologiests ‘Optimising Cereal Inputs: Its Scientific Basis’ Conference, p. 225-231. AAB, Wellesbourne, UK, AAB.

Weiss, E.A. (2000). Oilseed Crops. 2nd edition. Blackwell Science, Oxford, UK, Blackwell Science.

Wilkins, R.J. & Jones, R. (2000). Alternative home-grown protein sources for ruminants in the United Kingdom. Animal Feed Science and Technology, 85:, 23.

Wilkins, R.J., A. Hopkins, A., Dewhurst, R., Scholefield, D., Doyle, C. & Topp, K. (2001). Legume silages for animal production: increasing profits with forage legumes. IGER Report, North Wyke, UK.

Winters, A., Fychan, R., Davies, D.R., Lloyd, J., Merry, R.J. & Jones, R. (1999). Protein content of a range of ensiled legumes. pg. 131-132. In, P. Pauly, T. (ed.) Proceedings of the 12th International Silage Conference. p. 131-132. Uppsala, Sweden.

Appendix: A

Synopsis of alternative plant protein sources


Crude Protein
(g/kg-1 DM basis)


Aeschynomene spp., forage

150 - 250

Bird’s-foot trefoil, forage

198 - 210

Slow to establish

Brewers’ grain

170 - 320

Inconsistency of protein concentration between batches

Canola meal

430 - 450

Presence of glucosinolates, dealt with through the use of hexane and breeding

Cassava, forage


High tannin concentration could affect palatability

Centrosema acutifloium


Centrosema macrocarpum


Centrosema puescens


Less persistent than C. acutifloium

Cocoa residue


Contains theobromide - toxic to pigs, poultry, and horses, but can be fed to ruminants at 700mg/kg-1 feed

Coconut meal

220 - 230

Low in methionine, and cysteine

Cotton meal

410 - 440

Gossypol and aflatoxins can make it toxic to livestock especially monogastric

Crambe meal

400 - 600

Limited supply

Field beans, grain

260 - 300

Heat treatment required to deactivate several anti-nutritional factors

Groundnut meal

450 - 550

Aflatoxins possibly present and low in lysine and methionine

Jojoba meal

250 - 350

Detoxification through oil extraction needed, also low in methionine

Kudzu, forage

150 - 180

Low yielding, and does not tolerate close grazing

Kura clover, forage

146 - 208

Can cause bloat

Leucaena leucocephala


Minmosine is toxic to non-ruminants, limiting this species to ruminants

Linseed meal

350 - 380

May contain glucoside or linamirin, destroyed by high temperature processing

Lucerne, forage

129 - 324

Can cause bloat if grazed

Lupin, forage


Lupin, grain

320 - 400

Poor digestibility by monogastric animals

Macroptilium artopurpureum


Only tolerates light grazing pressure

Maize gluten meal

600 - 700

Lysine deficient

Niger meal


Low in lysine and threonine

Peas, forage

100 - 140

Peas, grain

250 - 260

Phyto haemagglutinis destroyed by heat treating

Quality protein maize, grain


Red clover, forage

140 - 280

Can cause bloat if grazed

Rhizoma peanut, forage

140 - 180

Establishment method, limits area sown to this crop

Safflower meal

200 - 240

Phytic acid makes the meal bitter and binds minerals

Sainfoin, forage

134 - 179

Difficult to maintain swards

Sericea lespedeza, forage

110 - 160

Very high condensed tannin content can interfere with protein digestion

Serradella, forage

200 - 270

Low yielding

Sesame meal

351 - 470

Low in lysine

Soybean meal

440 - 550

Trypsin inhibitor deactivated by heat processing

Soybean, forage


Soybean, roasted

370 - 440

Trypsin inhibitor deactivated by heat processing

Stylosanthes spp.


Badly suffers forom disease

Subterranean clover, forage

260 - 310

Sulla, forage


Poor persistence

Sunflower ensiled


Sunflower meal

360 - 400

Low in lysine

Previous Page Top of Page Next Page