It is argued that the growing discrepancy between the expanding world population and the earth's food producing capacity is primarily caused by competition between livestock and humans for cereal grains. The solution that is proposed is to develop alternative feeding systems for livestock using non-cereal, perennial, high-biomass producing crops that can be grown under rain-fed conditions on sloping lands not suitable for cereals. Examples are given of tropical crops that are several times more productive than maize and soya beans, are environmentally friendly and have multi-purpose use as food, feed and fuel.

The merits of these crops will be best appreciated in integrated farming systems in which: (i) the primary role of livestock is recycling crop residues and providing substrate for biodigesters and ponds for production of fish and water plants; and (ii) productivity is measured as the amount of solar energy that is harvested.

To respond to these complex needs, the professional agriculturalist of the future should be a generalist, trained in biology, ecology, socio-economics and communication.

Ways of conducting research, training and dissemination of information must be changed to reflect the above issues. Increasingly these activities will be conducted in the region and country where the results are to be applied. They will be farm- based and decentralized with responsibility for execution and reporting vested in the local community rather than the central government. But they will also be global, in terms of taking advantage of electronic exchange of information, which will enable researchers and farmers to be in daily communication with their colleagues in similar ecosystems in other countries and continents.

Three events will influence profoundly the role of livestock in farming systems in the next century. The first is the projected doubling of the world population, that will be concentrated in the less-developed countries. The second is the increasing imbalance between consumption and production of cereal grains. The third is the increasing aspirations of people in developing countries for the amenities taken for granted in the more- developed countries which will result in a massive increase in demand for energy in all its forms (Figure 1).

Future (irreversible) world events

A: Doubling of World Population (mostly in LDCs)

B: Increasing aspirations of people (especially in LDCs)

Will lead to pressure on:

Energy supply (to satisfy "B")

Food (cereal grain) to satisfy "A"

Opportunities for:

Livestock products ("B")

Renewable energy (biomass) ("B")

Nothing short of a series of major calamities (floods, earthquakes or uncontrollable outbreaks of disease) can halt the continuing increase in the world population at least over the next 50 years by which time there will be twice as many people to feed as at present.

By contrast the capacity of the earth's surface to produce more food appears to be close to saturation point (Brown and Kane 1994). The oceans are already over-fished and even to maintain present levels of production will be difficult unless more resolute steps are taken to preserve the natural breeding grounds. The demand for cereal grain already exceeds supply and by 2030 the projected world deficit is estimated to be of the order of 450 million tonnes (Figure 2) with production constrained by reduced availability of arable land and of water and by the need to reduce agrochemical inputs in the face of growing environmental degradation and the need to preserve biodiversity. In the face of shortages, grain prices will rise and their use to feed livestock will diminish.

The world presently derives some 60% of its energy from fossil fuels. The supplies of these are limited and at the present (1993) rate of consumption will last only 60 years; at the rate of consumption projected for the year 2005, the reserves of fossil fuel are likely to last less than 30 years (The Economist 1994). Inevitably, in time, prices will also begin to rise faster than the average rate of inflation. An equally worrying consequence of the projected increase in fossil fuel consumption is the emission of the greenhouse gas carbon dioxide, the rate of accumulation of which in the atmosphere is expected to double. Finally, economic development brings in its wake massive immigration from rural areas to cities, and eventually the social unrest caused by the loss of jobs as human power is replaced by mechanical and electrical devices.

The implications for agricultural production, especially in the less-developed tropical countries, are many and include:

• Cereal grains - which are the basis of livestock production in the more-developed countries - will face a competing demand to feed the increasing human population. This will result in higher prices.
• Increasing importance will be given to production and utilization of renewable forms of energy, especially biomass.
• Farming systems will be geared to practices which increase the sinks for the greenhouse gases carbon dioxide and methane.
• Natural renewable resources will have to be used more efficiently, farming systems will need to create more employment opportunities, especially for women and children, and the role of livestock in the system will be increasingly catalytic rather than a specialized activity.
• Mechanization will be increasingly based on animal rather than machine power.

This scenario will:

• Create opportunities for crops which: (i) have high photosynthetic efficiency; (ii) give rise to products that can replace cereal grains in the diet of livestock; (iii) are potential sources of renewable energy in the form of biomass; (iv) are environmentally friendly; (v) can be cultivated under rain-fed conditions and on sloping land; and (vi) give comparative advantage to small-scale farmers.
• Require new approaches to research, extension and training such that the present emphasis on individual commodities, yield maximization and uni-disciplinary specialization is replaced by integrated farming systems geared to optimum use of natural renewable resources, supported by services that are multi-disciplinary with maximum participation by the farm family.

The present dominant role of cereal grains in farming systems world-wide reflects the concentration of decision making in agricultural policy in the developed countries all of which are in temperate latitudes which provide comparative advantage to short season crops, especially cereals, as sources of useable biomass. By contrast the natural richness of the tropical latitudes is the climate and the limitations are the soils the fertility of which is dependent on recycling of nutrients from decaying biomass. These conditions favour perennial crops especially trees and tall-growing plants which generate large amounts of biomass a high proportion of which is recycled naturally to the soil. In the colonial era, with constraints on communication and availability of agrochemical inputs especially those derived from fossil fuel, these natural advantages were exploited and major advances were made in the development of many tropical crop plants (eg: coconut and oul palm, rubber, coffee, cocoa, sugar cane, bananas, spices). The discovery of oil and the development of agrochemicals, especially fertilizers, pesticides and herbicides, strengthened the comparative advantage of the developed (temperate) countries by making it possible to expand food production dramatically making available products such as sugar, edible oil and (synthetic) rubber, which previously were imported from the tropics. At the same time, improvements in communication favoured still more the developed countries, with the result that the professional human resource base in tropical countries became increasingly influenced by the farming systems and technologies being developed in the temperate countries. The outcome of both these trends has been:

• The introduction and promotion of fossil fuel-based agricultural systems, in most cases in conflict rather than in harmony with tropical ecosystems.
• A belief that intensification of livestock production depends exclusively on the use of cereal grain and protein-rich meals (the "maize and soya bean" philosophy)

The above thinking is exemplified in the otherwise excellent analysis made by Brown and Kane (1994) of the risks and challenges inherent in the conflict between the needs of the expanding world population and the capacity of the earth's surface to feed them. The predictions of a world grain deficit of some 450 million tonnes are predicated on the present agricultural practices whereby 40% of the grain is fed to livestock. If all the world's future supplies of cereal grain are reserved for people there will be enough food (240 kg/person according to Brown and Kane) to feed the 10 billion inhabitants which is the point at which the world population is expected to stabilise. The challenge is thus not so much how to expand cereal grain production but rather to develop other cropping systems that use natural resources more efficiently.

This challenge, coupled with the knowledge that fossil-fuel based farming is neither sustainable nor capable of providing the required resources for the increasing world population, is an opportunity for tropical agriculturalists to assume the leadership in developing tropical ecosystems that offer viable alternatives which are sustainable and have the potential to satisfy the increased demand for food, feed and fuel.

The first step is to select those tropical ecosystems that are both productive and sustainable under rain-fed conditions on sloping soils not suitable for rice production. The second step is to integrate crops and livestock with biodigesters and ponds (Figure 3). Five crops already show excellent promise in their capacity to provide the energy component of the diet - for people and livestock. These are:

• Sugar cane
• African oil palm
• The sugar palm
• The Coconut palm
• Cassava

All these crops, with the exception of cassava, are perennial and can be grown continuously in the same soil provided the fallen dry leaves are allowed to decompose on the soil surface. Cassava is an annual crop which must be replanted every 6 to12 months. Its potential to recycle dead leaves is neligible and it should be grown in rotation with fertility-restoring crops such as sugar cane. Its advantage is its capacity to grow on poor soils with minimal rainfall.

As protein sources there is a vast potential in certain water plants (Azolla and Duckweed) and in multi-purpose trees (Gliricidia sepium, Trichanthera gigantia and Leucaena leucocephala).

Other candidates, to supply energy and protein, will certainly appear once research into tropical ecosystems is given the same support as has been given in the past to temperate crops. Some comparisons of yields will put this argument into perspective.

The productivity of maize in the USA and of wheat in the UK, countries which are the world leaders in these crops, is respectively 6 and 8 tonnes grain/ha with similar amounts of biomass in the straw/stover residue (Brown and Kane 1994). These yields are obtained on fertile soils with adequate moisture.

Contrast these data with recent reports from tropical countries.

• Sugar cane on highly fertile land in the Cauca valley of Colombia produces 150 tonnes/ha of fresh biomass containing 18 tonnes of sugar and a similar amount of fibrous residue (ASOCANA 1994). On sloping hillsides in the Andean foothills, with minimum inputs, yields are of the order of 60 tonnes stalks/ha producing 6 tonnes of crude brown sugar (panela).
• The African Oil Palm in Malaysia averages 20 tonnes of fruit/ha containing 4 tonnes oil, equivalent in energy value to 9 tonnes of grain; the theoretical peak yield is estimated to be 45 tonnes/ha annually (Gray 1995).
• The Sugar Palm in Cambodia (under rain-fed conditions) produces on average 750 litres of juice with 16% of sugars in a 150 day harvest season while occupying an area of less than 50 m² per tree (Borin Knieu and Preston 1995). This is equivalent to an annual yield of 24 tonnes of sugar/ha.
• In Sumatra, the coconut palm flower is manipulated to yield juice at the rate of 3 litres/day, containing 450 g sugar, throughout the year (Levang 1988). With a population of 120 trees/ha this is a yield of 19 tonnes sugar/ha.
• Cassava produces on average 20 tonnes roots/ha in growing seasons of 6 to 12 months, equivalent to 8 tonnes cereal grain.

The market leader in protein production in the developed countries is soya bean, which in the USA - the country with the most advanced technology - yields on average 2 tonnes/ha. This is about 750 kg of protein/ha.

In the tropics:

• Duckweed grown in water fertilized with biodigester effluent produces on average 300 tonnes of fresh biomass containing 5 tonnes of protein (Rodriguez and Preston 1995). Yields as high as 10 tonnes protein/ha have been reported. The Azolla water fern has a similar potential (Rodriguez and Cuellar 1994).
• Gliricidia sepium averages 60 tonnes/ha of fresh biomass containing 3 tonnes of protein (Gomez et al 1990). Trichanthera gigantea yields a similar amount (Gomez and Murgueitio 1991)

The above comparisons show that perennial plants in tropical ecosystems out yield by factors of 2-3 the equivalent short season crops grown in temperate latitudes.

Future criteria for sustainable use of natural renewable resources must be based on the output of the overall farming system in accordance with the needs of the farm family and the market, and not simply the results for individual elements such as crop yield, live weight gain or feed conversion. An appropriate indicator would be the capacity of the farming system, and of the community of which the farm is a part, to convert solar energy into useful products for the family and the market. An example of the application of this approach is to evaluate pig production managed as a specialized (commodity) enterprise or as a component of an integrated farming system. To obtain a high growth rate and a good feed conversion in growing pigs (the commodity approach) requires a diet of high digestibility which means less manure for the biodigester, less gas for cooking and less fertilizer for the fish pond (where pig production is part of the farming system). For the latter situation, it may be better to feed a diet of lower digestibility (especially if such a diet can be produced on the farm), which probably will result in lower rates of live weight gain and poorer feed conversion but may contribute more to the total system.

Re-training the professional agriculturist

NOT in "developed" countries.

Their Agriculture is not sustainable (oil subsidies!!)

On farms in villages, learning by doing

The "farmer" as "professor"

Knowledge of the biological principles underlying efficient

use of natural resources can be obtained by E-mail

Provides immediate access to experienced scientists and to

literature, especially "grey" (relevant) literature

There are more computerized agricultural scientific

journals in developing than in developed countries

To respond adequately to the needs of the farming system approach, the professional agriculturalist must be first and foremost a generalist - trained in biology, ecology, socio- economics and communication - rather than in crop or animal science. In this respect, the student can learn as much from farmers - who are essentially practising biologists - as from professional scientists. Communication is especially important because to have ready access to knowledge obviates the need to memorize information and to sit in classrooms listening to often irrelevant lectures. The electronic highway, as exemplified by the Internet and the scores of local e-mail networks, has democratized access to information making it available to anyone with a computer, modem and telephone - essential tools of the future professional agriculturist.

Ways of conducting research, training professionals and disseminating extension messages, must be changed to reflect the issues discussed here. Increasingly these activities will be conducted in the region and country where the results are to be applied. They will be farm-based and decentralized with responsibility for execution and reporting vested in the local community rather than the central government. But they will also be global, in terms of taking advantage of electronic exchange of information, which will enable researchers and farmers in a village in one continent to be in daily communication with their colleagues working in similar ecosystems the other side of the world.