The Genetics and Exploitation of Heterosis in Crops: An International Symposium (1997)

Heterosis: Feeding People and Protecting Natural Resources
D.N. Duvick
Department of Agronomy, Iowa State University, Ames, IA 50011, USA.

Hybrid vigor, the increase in size or rate of growth of offspring over parents, has been carefully examined, given the name "heterosis", and utilized on a large scale for production of selected hybrids of plants and animals for the past 75 years. Field crops such as maize, sorghum and sunflower are now produced as hybrids in all of the industrialized world; they are also are grown increasingly as hybrids in developing countries. Hybrid rice is grown extensively in China and is now introduced into India (Virmani 1994). Many commercial vegetable and flower crops are grown almost entirely as hybrids. Heterosis is credited for large increases in production per unit area and therefore for sparing large amounts of land for other uses such as environmentally benign nature preserves (Table 1). Examination of the historical record shows, however, that heterosis per se may have been only a catalyst rather than the primary cause of annually increasing yields for hybrid crops, worldwide.

Hybrid maize was first bred and produced in the USA. The first hybrids yielded about 15% more than the better open pollinated varieties. Starting in the 1930s the area planted to hybrid maize started to increase rapidly and after 15 years 95% of the land in the US Corn Belt was planted to hybrid maize. Coincidentally, with the rise in plantings of hybrid maize, on-farm maize yields began to climb. They continued to climb after US maize plantings were essentially 100% hybrid (about 1965) and they are still rising (Fig. 1). Several studies have shown that about 50% of the yield gains since the 1930s is due to changes in management, such as increases in nitrogen fertilizer and higher plant densities, while the other 50% is due to changes in maize genotype (Duvlck 1992). Inbred yields have increased over the sears at nearly the same rate as the yields of their F1 hybrids (Fig. 2), therefore heterosis has not increased, in fact it has decreased if calculated as percent advantage of hybrid over parents (Duvick 1984). Experiments demonstrate that yield gains in hybrid maize are due primarily to improvements in tolerance to abiotic and biotic stress, and that the improvements occurred in parental inbreds as well as in their hybrid progeny

Grain sorghum has been grown as hybrids for about 40 years (Doggett 1988) and sunflower for about 20 years (Miller 1987). Measurements of sequential changes in heterosis have not been made for either of these crops. Breeder observations indicate that for both crops modern inbreds are not uniformly higher yielding than inbred parents of the early years, therefore heterosis is somewhat greater since new hybrids outyield the older ones. However, in both crops some of the best new inbred parents are clearly more vigorous and higher yielding than their predecessors. As with maize, gains in hybrid performance of both crops are due primarily to improvements in traits that confer stability of performance; e.g., in disease and insect resistance, drought tolerance, staygreen (for sorghum), and standability. Parents as well as hybrids have acquired these improvements; they are not the unique product of heterosis. Thus, changes in traits that could just as well have been improved in the open pollinated varieties (but with more difficulty) are responsible for improved hybrid performance in sorghum and sunflower, as well as for maize.

Heterosis has been responsible for much of the annual yield gains in maize, sorghum, and sunflower, but its effect was indirect. Its indirect benefits include: 1) Precise genotype identification and multiplication. Instead of a random collection of hybrid/inbred plants in an open pollinated variety, the most superior hybrid combinations can be identified and reproduced at will in unlimited quantity. 2) Breeders of hybrid crops can react faster and with more options to meet changing times and changing demands. New hybrids with needed new traits can be made and tested within one or two seasons, given a broad-based pool of inbred lines. 3) Farmers can easily identify hybrids as compared to their open pollinated varieties; they expect more from hybrids, they are more likely to provide extra inputs, and they keep constant pressure on breeders to make further improvements in hybrid performance. 4) The prospect of annual seed sales at profitable prices attracts private capital to hybrid breeding and sales. Therefore maize breeding and associated seed production and distribution technologies are doubly supported, by both public and private funds.


Table 1. Estimates of the current annual global contributions of hybridization to production of maize, sorghum, sunflower, and rice.

Crop % area
planted to
% hybrid
% annual
added yield
Annual added
(million t)
Annual land
(million ha)
Maize 65 15 10 52 13
Sorghum 48 40 19 13 9
Sunflower 60 50 30 7 6
Rice 12 30 4 15 6
*Estimated gain in yield of hybrids over superior open pollinated varieties at time of hybrid introduction.
Figure 1. USA maize yields 1901-1996 Hybrid maize was introduced in about 1930 and was used on 100% of USA maize plantings by about 1965. (USDA NASS. "Agricultural Statistics"). Figure 2. Yields of maize single crosses (SX) and the means of their parent inbred. (MP); widely used pedigrees in Iowa (USA) in each decade. 1930s through 1960s. (Duvick, unpublished data).