In a context of increasing maize demand in the developing world, appropriate policies should encourage a greater supply of maize. In the USA, maize yields have increased dramatically, due to the adoption of hybrid varieties, use of chemical inputs, and increased input efficiency in different periods respectively. Hybrids assured private involvement in the US maize seed industry. If Latin America is to follow in the footsteps of the USA, it should encourage the development of newer hybrids and improved varieties adapted to different agro­climatic conditions. Private and public sectors could be complementary in this strategy.

Maize was the first cereal crop to undergo rapid and widespread technological transformation, as the history of hybrid maize in the USA shows. The experience gained with maize deeply influenced other crops. Even the birth of the Green Revolution can be traced to it. Henry Wallace, the initiator of the Rockefeller Foundation's agricultural research programme in Mexico in the 1940s was the founder of one of the first and most successful hybrid­maize seed companies in the USA. When the Rockefeller Foundation shifted its attention to India, the initial emphasis was also on hybrid maize.
The success of technological change in maize production on a global scale, reflected in continuing yield increases, is heavily based on the use of so called high­yielding varieties (HYVs). Notwithstanding this success, many important maize­producing areas of developing countries remain largely untouched by these new technologies. According to the International Maize and Wheat Improvement Centre (CIMMYT, by its Spanish acronym), only 20 per cent of Mexico's, roughly a third of Central America's and only 15 per cent Colombia's total maize production area is sown with HYVs varieties (both open­pollinated varieties and hybrids). Two types of maize agriculture could roughly be distinguished: firstly, the commercial, high­input production, found in the US Corn Belt and in several significant pockets in developing countries such as Argentina and Brazil. Secondly, the small­scale, semi­subsistence production exemplified by most of the maize producers in developing countries.
Maize yields in the USA are close to six tonnes per hectare per year. Small­scale farmers in developing countries obtain less than two tonnes per hectare per year. In spite of this difference, the two economies are increasingly linked through world trade, given the economic liberalization implemented by many developing countries in the 1980s and accelerated in the 1990s. The drastic reduction of import barriers implemented by Latin American countries in the last decade place the local farmers in direct competition with the high yield producers of the exporting countries, mainly the USA and Argentina. Therefore, technological change in maize production has to be viewed within a global perspective, taking into account not only the potential for change in one country, but also the trends in maize production technology in exporting and importing countries.

Trends in US commercial maize production
The story of commercial maize production is best exemplified by the case of the USA, by far the world's largest producer and exporter of maize. Until the 1930s, most of the increase in its maize production came from expansion of the crop area, with only modest changes in yields. Beginning in the 1930s, maize production entered a new phase characterized by yield increases obtained through hybridization. These hybrid maize varieties rapidly spread in the main Corn Belt, and then entered other areas of the USA. The first hybrids were quickly replaced by newer generations resulting in an average growth rate in yields of 2.7 per cent per year over the period 1930­55 (which is somewhat below the rapid growth of rice and wheat yields in Asia achieved during the Green Revolution). It is important to recognize that the main source of growth of yields during this period was not the switch to hybrid maize per se, but the genetic gains realized through the periodic replacement of older hybrids by newer hybrids.
From the mid 1950s onwards, use of chemical fertilizers and pesticides rapidly expanded. By the 1980s, the nitrogen use averaged about 150 kg/ha compared with less than 10 kg/ha in 1955. In addition, herbicides had almost universally been adopted and insecticide use had increased substantially. This phase of input intensification was promoted by declining prices of these inputs in the post­war years, combined with the development of newer hybrids with a much greater response to increased doses of nitrogen and higher plant densities. Thus, in the period 1955­80, maize yields in the USA continued to increase over 3 per cent annually.
In the 1980s, the growth rate of maize yields in the USA slowed due to diminishing returns to the use of increased doses of inputs and the absence of any major technical breakthroughs. At the same time, concerns about the environmental consequences of the use of high levels of chemical inputs increased. This led to a switch in strategy from input intensification to input efficiency, along with conservation of the natural resource base, especially land and water. For example, through an extensive campaign of research and extension, the US state Iowa has reduced the average dose of nitrogenous fertilizer on maize by some 25 kg/ha or 15 per cent in the period 1985­91, without any effects on yields. It also has been successful in reducing the use of pesticides in maize production.

The general pattern of technical change in maize production in the USA holds for most other countries and regions where maize is produced commercially. For example, hybrid maize technology and seed was exported to Europe in the post­war period, as well as to Argentina, Brazil and Chile. These areas have also experienced a period of input intensification and are now increasingly concerned about enhancing input efficiency and resource conservation.

Hybrids in Latin America
Argentina and Brazil were the first Latin American countries that mastered hybrid technologies. As early as 1923, a US researcher was reported to be working on the subject in Argentina. By 1945, two locally developed double­hybrids were registered: The Santa Fe 2 and Santa Fe 3. By the late 1940s, the private sector was increasingly involved in maize hybrid seed. For example, Cargill, USA, created a seed production facility in Argentina. Later, the University of Buenos Aires registered several hybrid varieties. The lines used to produce these hybrids were transferred to the private seed industry and were used for a substantial time as the genetic base of private hybrids.
In the 1960s, a hybrid­seed industry had already consolidated in Argentina, partly due to the large market it served. By acquiring national companies, transnational corporations have been extensively involved. A permanent stream of newly developed hybrids has characterized the industry from that time until the present.

Brazilian seed industry also goes back to the early decades of the century. During the late 1920s, a maize breeding unit was established at the Instituto Agronomico de Campinas (IAC), in the State of Sao Paulo. Some of its researchers studied genetics in US universities and, upon their return to Brazil, established a maize breeding programme. In the late 1930s and early 1940s, the combination of Brazilian with other Latin American germplasm led to superior hybrid varieties, such as the Cateto and Azteca. In 1946, the first double­crossed hybrid was officially released by the IAC.
In 1945, the main two breeders of the University of Viscosa started a new company, AGROCERES, with financial support from the Rockefeller Foundation. This company is now the most important seed company in Brazil, and one of the largest agribusiness conglomerates in the country. By the decade of the 1960s, the maize seed industry in Brazil was firmly established, with AGROCERES holding a dominant share of the market. During this decade, two major transnational corporations, Cargill and Pioneer Hi­Bred, USA, entered the market. EMBRAPA, the national agricultural research institute, had also released hybrids.

In these two countries, the maize economy can be characterized as a commercial one, with high­input production systems and varieties that can take advantage of these inputs. In 1990, Argentinean maize production was based 100 per cent on improved varieties and hybrids. In Brazil, this figure was much lower (57 per cent), reflecting the presence of a small­scale subsector in the North of the country. Most breeding efforts have taken place in, and has been directed to the relatively advanced South and Central areas of the Brazil.

Mexico's maize production has followed a different path. The typical farm in Mexico is much smaller than in Brazil or Argentina. The Mexican government supplied these farmers with seeds through a specialized corporation, Productora Nacional de Semillas (PRONASE). This corporation had the mandate to produce and distribute commercial seed of all varieties created by INIA, the national agricultural research institute. In fact, PRONASE's distribution of these varieties was a quasi­monopoly. Private sector involvement in the seed industry developed slowly, constrained by lack of access to INIA's maize germplasm.
Adoption of improved varieties and hybrids in Mexico's main staple crop has been very low, even by Latin American standards. One of the reasons could be the lack of private sector involvement in the seed industry, together with the predominance of small­scale farmers. Additionally, maize is grown under many different agro­ecological conditions. Because of its importance as food, many different traditional varieties exist, adapted to the different conditions and specific final uses. Recently, the seed industry regulations in Mexico have been completely overhauled, resulting in PRONASE loosing its monopoly privileges. As a consequence, a very active private seed industry has sprung­up in the last two years.

The dynamics of seed improvement
In the rest of Latin America, and in the developing world in general, the adoption of HYVs by small­scale farmers has been slow. The reasons for this are many, such as for example, low farmer's expectations of gains in yields brought by HYVs, and higher costs of the seed and additional inputs. Even if hybrids or improved open­pollinated varieties look profitable, some producers, especially small­scale farmers, find it difficult to adopt them, simply because of a lack of access to credit. Nevertheless, experiences in China, Venezuela and Zimbabwe, where hybrid seed is relatively cheap, suggest a high potential of HYVs to be adopted by different types of producers.
As the seed industry develops and matures, and most farmers increasingly adopt hybrids, seed prices increase, encouraging private involvement in maize breeding and propagation. The experience of industrialized countries, where the ratio seed/grain price doubles the ratio typical of developing countries, suggests that seed price sensitiveness among farmers is a function of access to resources, and the farmers' positive view on the yield response of hybrid seeds.
For the increase of maize production by small­scale farmers, the yield difference between open­pollinated varieties and hybrids is not so relevant as the difference between modern varieties (open­pollinated varieties or hybrids) and non­improved ones. But, the difference between open­pollinated varieties and hybrids is very important in terms of who will invest how much in maize improvement. In other words, an important question is how to guarantee a continuous production of improved technology, both in terms of new germplasm and its associated inputs and agronomic practices. The inborn protection that hybrids have compared to open­pollinated varieties makes them very attractive for the private plant­breeding sector. In fact, the development of the hybrid technology permitted the creation of a private seed industry.
On the other hand, it is cheaper to produce open­pollinated variety seed than hybrid seed. This motivated many governments in Latin America during the 1960s and 1970s to generate open­pollinated varieties. For example, nearly 70 per cent of maize varieties released by Central American and Mexican national agricultural research systems (NARS) during the period 1966­80 were open­pollinated. More recently, there has been a shift toward more hybrid release. During the 1980s, nearly 50 per cent of the released materials in Central America were hybrids. This change of priorities might reflect the fact that, from the point of view of more commercial farmers, hybrids can be more profitable than open­pollinated varieties.

Private and public sector
Private sector investment in research has been a major driving force in the increase in productivity of commercial maize production. In the USA, it has been estimated by CIMMYT that this investment increased from US$ 8 million (rate 1990) in 1955 to US$ 110 million in 1990. In terms of its share of output value, it has grown from less than 0.1 to over 0.5 per cent of the value of total maize production in the USA. It has been estimated that in the mid­1970s public sector maize research expenditures in developing countries on all aspects of maize production was 0.23 per cent of the value of maize production in developing countries, or about US$ 70 million (rate 1990). CIMMYT estimates that twenty years (1990) later about US$ 85 million was spent on maize improvement in developing countries, and that the private sector accounted for about one quarter of these expenditures. Surprisingly, there are about 9 breeders per million tonnes of maize produced in developing countries, compared to 4 in the USA, with most of this difference being accounted by the public sector.

We would like to raise two questions about maize research in developing countries. First, should the private sector assume a greater share of investment in maize improvement research? This trend is already evident in countries with large markets for hybrid seeds, such as Brazil, Mexico and Zimbabwe.
Second, has the public sector found an appropriate balance in its investment between maize improvement research versus crop and resource management research? The total investment in maize improvement is quite high, while the private sector is increasing its share. Private industry, however, is unlikely to make profitable investments in crop and resource management research. Therefore, there is a strong case for the public sector to give more emphasis to research on crop and resource management. In early stages of the seed industry development, an active government intervention could be justified. Later, co­operative arrangements between public and private sector could become more important, as exemplified by the cases of NARS in Mexico, Brazil and Argentina. In these countries the NARS produce basic seed, which is reproduced by private entrepreneurs.

Changing research priorities towards sustainability
Because the use of seed­fertilizer technology in the small­scale farmer sector is still very incomplete, efforts to enter the next stage of technical change emphasizing cost reduction, input efficiency and conservation of natural resources have not received much attention to date. Recent experiences, however, suggest that there are substantial opportunities in the small­scale farm sector to increase productivity and sustain the resource base through enhanced input efficiency. Consequently, the adoption of more environmentally­friendly technologies, as applied in for example the USA, can also be financially attractive for farmers in developing countries.
In La Fraylesca, Mexico, farmers apply over 150 kg/ha to achieve yields of only a little over 2 tonnes/ha. Acidification of soil has been identified as the major cause for this inefficiency, which could be alleviated through liming. Likewise, in other tropical maize producing areas of Mexico on­farm research has identified a range of factors which could be profitably introduced to increase maize productivity among small­scale farmers. One important point to notice is that the opportunities to increase productivity are very site­specific, and require a well developed local adaptive research and extension system.
Within a more medium term perspective, site­specific constrains on maize yields in developing countries (both biotic and abiotic) could be overcome by developing more site­specific germplasm. Biotechnology could be a factor in this challenge since costly conventional breeding is only economic when it produces materials for relatively large markets, i.e. extensive relatively homogeneous agro­ecological conditions. The development of germplasm for smaller markets, real niches, therefore requires a significant reduction in the costs of plant breeding, to which biotechnology might contribute in the medium term. Similarly, enhanced input use efficiency is linked to biotechnology developments.

Recent developments in maize research in industrialized countries have centred on biotechnology and resource­conservation strategies. Although the rate of progress in maize has been much slower than anticipated, there is no doubt that maize seed sales by the year 2000 will incorporate biotechnology applications, like hybrids with genetic resistance to herbicides and pests. The utilization of molecular techniques to develop improved varieties for small­scale farmers is only very recently being initiated by public sector programmes, including CIMMYT.
Whether it will be possible to implement new policies towards newer hybrids and improved varieties adapted to different agro­climatic conditions in developing countries in the midst of fiscal constraints derived from structural adjustment is still uncertain. Evidence from Mexico, Brazil and Argentina suggests that new relationships between the private and public sector, where the latter concentrates on R&D and the former on production and marketing, could be an opportunity.
Walter Jaffé/Miguel Rojas (IICA)

Sources
Derek Byerlee and Miguel A. Lopez­Pereira (1993), Technical Change in Maize: A global perspective. Presented at the First International Maize Symposium, March 16­19, 1993, Guadalajara, Mexico.

Miguel A. Lopez­Pereira and Joao Carlos Garcia, The Maize Seed Industries of Brazil and Mexico: Past performance and future prospects. Forthcoming.

Miguel A. Lopez­Pereira and Alejandro Espinosa Calderon (1993), Análisis Económico de la Producción y Uso de Semilla Mejorada de Maíz: El caso de Mexico. Paper presented at the PCCMCA, March 1993.

Eduardo Jacobs and Marta Gutierrez (1984), La Industria de Semillas en Argentina. Buenos Aires: Centro de Investigaciones Sociales sobre el Estado y la Administración.