We know the consensus
We know the most likely agriculture scenario
What we know: future yields will involve higher fertilizer use
What we guess: the future of fertilizers
We don't know much about environment trade-offs
We guess about other nutrient sources
We also don't know much about soil biology
We know about crop improvements
We know the poor progress in achieving the WFS goal
We know the WFS implications for fertilizer use
We know potential gains are considerable

We know the consensus

There appears to be some consensus in the research community about the likely future path of global agricultural production and natural resource use (IFPRI, 1995; NAS, 1998) . Aspects of this consensus can be succinctly summarized as follows: growing world population and per capita incomes will likely require more intensive agricultural crop production. Higher yields will, in turn, increase the demand for, amogst others, nutrient inputs. Furthermore, future agricultural cropping patterns will reflect shifts in diets (e.g. increased meat consumption in developing countries). Greater opportunities for agricultural trade may also lead to regional shifts in world crop production. At the same time there will likely be economic and environmental incentives to improve the efficiency of fertilizer use over current levels in all countries, especially in the developed countries.

We know the most likely agriculture scenario

By 2030 world population will be approaching it’s peak and 5 000 years of arable land expansion will be almost over. In 1960 two third of all people resided in rural areas, in 2030 almost the same proportion will be urban dwellers. Growth in food and agricultural production will be largely dependent on land use intensification and on technologies that are increasingly focused on minimising impacts on the environment and raising resource use efficiency (particularly of water) rather than on improving yields. The overall picture coming from the FAO analysis is for agricultural production to become:

• primarily demand driven rather than supply driven;
• for supply to become even more technology dependent though with some of the technologies being based on the application of science to raising the productivity of natural processes rather than of external inputs; and
• for agriculture to exert declining pressures on natural resources and the environment.

Agricultural intensification further increases flows of plant nutrients to crops and results in a higher nutrient uptake. The depletion of nutrient stocks in the soil, which is occurring in many developing countries, is a major but often hidden form of land degradation, making agricultural production unsustainable. For the farmer, nutrient losses are, at least in the long run, a financial loss.

What we know: future yields will involve higher fertilizer use

Increased harvested area (rainfed and irrigated) and in particular higher yields provide the required increase in production under the “business as usual scenario”. Because fertilizer use in Africa is low, farmers will achieve higher aggregate yields at national level only when fertilizer use becomes more profitable. This implies that increased demand for food becomes effective, that increased food supply originates primarily from domestic production and that improved fertilizer response, for which investment in soil fertility is a prerequisite, materialises. Only 17 million tonnes of mineral fertilizer were applied in 1950, four times more than in 1900 but eight times less than today. From the late 1940s to the end of the 1990s, average wheat yields increased from 1 100 to more than 2 600 kg/ha in the United States, while mineral fertilizer use rose from 20 to 120 kg/ha of arable land. In France, wheat yields increased from 1 800 to 7 100 kg/ha for 45 and 250 kg of fertilizer, respectively. Nowadays, on the rich loamy soil of north-western Europe, wheat and maize yields sometimes exceed 10 000 kg/ha, with fertilizer applications of about 200 kg/ha of nitrogen, 50 kg/ha of phosphate and 50 kg/ha of potassium.

What we guess: the future of fertilizers

Plant nutrients are key in the transformation of agriculture. They are a pre-requisite for production and a source of major environmental trade-offs. The ratio of mineral fertilizer use to other nutrient sources in different parts of the world varies widely within regions and even communities, according to the intensity of production, the crop or crop-livestock system, the climatic conditions, access to markets, infrastructure, services and income. It also depends on the level of knowledge and application of a wide range of organic and biological sources of nutrients, including nitrogen fixation by legumes and non-legumes, use of crop residues, animal wastes, as well as agro-industrial products, compost and municipal wastes. Global crop production in the early 1960s removed some 70 million tonnes plant nutrients which constitute 37% of the global nutrient cycle and will slightly increase to approximately 40% in 2015. The relative share of mineral fertilizers considering all sources of global nutrient inputs available for crop production, however, is projected to increase from 43% in 1960 to 84% in 2015. The relative importance of mineral fertilizers is thus expected to almost double in a time span of half a century. The contribution of non-mineral plant nutrient sources would slightly decline from 40 million tonnes nutrients in 1960 to some 30 million tonnes nutrients in 2015. The most likely scenario unfolding suggests food production to expand in response to increased demand with modest increases in mineral fertilizer use over the next 30 years (at 0.9% per year) though with considerable regional differences: i.e. almost 2% in sub-Saharan Africa and less than 1% in East Asia.

We don't know much about environment trade-offs

In parts of Africa mineral fertilizers are irreplaceable given the low supply of organic manure and biological sources of plant nutrients. In other areas, however, they are replaceable to a limited extent if land and labour can be substituted at a price consumer markets will accept. There are a range of substitutions and trade-offs with high input systems reducing the need for deforestation to produce more arable land. And herbicides permitting the use of conservation tillage systems with environmental benefits arising from greater plant nutrient and water use efficiency, lower soil erosion, and increased carbon sequestration. As with irrigation water, there are a number of institutionally and technologically feasible possibilities for raising fertilizer use efficiency.

We guess about other nutrient sources

Improving nutrient use efficiency requires a combination of local/farmer and scientific knowledge to continuously adapt and improve crop, soil, and water and livestock management practices taking into account the social, economic and environmental context. Plant nutrients in human excreta will be the largest contributor to nutrient outflows and exceed the sum of outflow from all other sources such as crop production, P fixation, erosion, NH 3-NO x emissions and leaching. The US EPA estimates that 20% of all sludge is treated in developed countries and 5% in developing countries. 20-50% of the treated sludge in developed countries is applied on agricultural land and consequently it is safe to guess that less than 1% is recovered on a global scale. Accelerated urbanisation in developing countries will hold the potential to achieve higher nutrient recovery from human excreta for crop production as the cost of sludge treatment, and in particular energy cost, reduces over time.

We also don't know much about soil biology

Protecting and improving the soil also makes good business sense. Research has shown that on relatively good soils initial nutrient recovery was only about 30 percent, but after 4 to 7 years of soil improvement, nutrient use efficiency increased two to three times. Without soil improvement, in fact, the capture of nutrients is only about 35 percent for nitrogen and 15 percent for phosphorus that is approximately half of rates typical elsewhere. This is particularly important in Africa where roughly twice as many nutrients are said to be lost compared to other regions, so that the majority of available nutrients are not utilised by crops. Above all, the natural process of biological nitrogen fixation (BNF) constitutes an important potential source of nitrogen for crop growth and protein production in many soils and ecosystems. It has recently been estimated that 40-48 million tonnes N per year is biologically fixed in agricultural crops and fields. In comparison, 83 million tonnes per year are currently fixed industrially for the production of fertilizer. Nonetheless, besides best management practices to secure a reasonable yield, further increasing the efficiency and productivity of legume - Rhizobium associations is relatively complex. A major limitation of legume rotations is that legume roots are low in N thus carry over is only sufficient for moderate production of a following crop. There is a need to develop better cereal-legume rotations that cause less acidification (and Al toxicity) and to better understand interactions (P, trace elements, etc) and efficiency of fertilization and fixation, taking into account soil biological management.

We know about crop improvements

The impact of biotechnology and genomics on fertilizer use efficiency are difficult to assess at this stage. Yield genes that have an improved capacity to assimilate nitrogen reduce environmental pressures by decreasing nitrogen leaching into groundwater. This means that sustained nitrogen losses in non-improved material are converted into higher yields by genetically improved material. Nutrient losses become lower and nutrient use efficiency thus increases. Farmers produce higher yields with proportionally lower incremental nutrient inputs. An additional fertilizer requirement is induced when nutrient losses are below the quantity of nutrients required to support potential yield limits. Consequently, fertilizer demand will increase. Biological nitrogen fixation by non-legumes through genetic manipulation is unlikely to emerge any day soon, the annual ASA meeting is conspicuously silent on the topic. The development of stable material with the desired characteristics for commercial dissemination requires some 20 years of research. At present applications aim at produce quality improvements rather than yields.

We know the poor progress in achieving the WFS goal

The latest FAO estimate of undernourished people in the world puts the figure at 815 million in the period 1997-99. But to achieve the World Food Summit goal of halving the number of undernourished in developing countries by 2015, the average annual decrease required is 22 million - well above the level of performance in the early post summit years. FAO has estimated that total annual gross investment in agriculture in developing countries needed to reach the WFS target of halving the number of hungry by 2015 would be about US$ 180 billion.

We know the WFS implications for fertilizer use

The total nutrient demand must be met from various organic and mineral sources, of which mineral fertilizer has currently the largest share (40%). Consumption of mineral N fertilizer alone was 78 million t.yr -1 in 1995-7 and, assuming a constant relationship between fertilizer use and crop production, is estimated to reach 96 million t.yr -1 to meet increased food demand for the year 2030. Achieving WFS food production targets would involve higher fertilizer application. Fertilizer consumption in developing countries would need to increase by 8 % in 2015 compared to a "business as usual" scenario. Such enhanced fertilizer use would be in particular critical in South East Asia, which accounts for two thirds of the total increment. Although such an increase is modest from a global perspective, some countries would need to dramatically increase fertilizer use if the WFS production target is to be achieved. In Asia, DPR Laos, The Philippines, and China and in Africa, Uganda, Côte d'Ivoire and Ghana would need to achieve substantial higher fertilizer applications to attain higher yields or, alternatively, resort to higher food imports. Will present farmers who currently do not use fertilizers become fertilizer users? We do know fertilizer use profitability to be the primary factor in the equation that governs fertilizer use adoption. Crop fertilizer response is the only factor that farmers can manage to some extent, low or declining commodity prices and constant or slightly declining fertilizer prices are determined by markets. Improvement in crop fertilizer use response is therefore an imperative for the food insecure small farmers in many developing countries.

We know potential gains are considerable

The scope for raising fertilizer productivity is substantial. Furthermore, organic materials are key to maintaining and restoring soil nutrient exchange capacity. An increase in mineral N use efficiency in rice production from 40 % recovery to 50 % in 2015 in Asia for instance has the potential to increase total rice output with 40 million tonnes or 3 % of global production. Farmers would save some $500 million per year and investment in urea manufacturing capacity would be $1.5 billion lower. Achieving such potential productivity gains would depend, however, on investments, research and technology transfer as well as policy support and incentives for the adoption of better management practices. To sum up, the world needs more mineral fertilizers and increased nutrient use efficiency, especially in intensive crop production systems. We can console ourselves with the knowledge that much of what we know still needs to be implemented to a large extent despite future uncertainties in scientific developements.