Some writers, noting the rapid turnover of popular cultivars, have suggested that modern growers substitute temporal for cross-sectional diversity. This might seem consistent with continued reliance on new genetic material from the centres of diversity, but the rapid turnover of varieties does not imply that there are continued large-scale inflows of germplasm.
In rice, Evenson and Gollin (1994) show that the amount of new germplasm introduced in IRRI releases seems to have declined in recent years, as these releases share much of the germplasm of previous releases. Importantly, all are reported to incorporate the same semi-dwarfism locus sd-1, and the Cina cytoplasm is still pervasive (National Research Council, 1993, p. 76). This does not mean that exotic germplasm has been completely ignored. IRRI breeders have effectively incorporated successive genes for pest and disease resistance from exotic germplasm; the complexity of this enterprise is illustrated in the account of Plucknett et al. (1987, Ch. 9) of the development of IR36. These genes enter as rare traits via successive backcrossing, so that the effective expansion of the germplasm is rather modest.
In maize, the major corn cultivars all trace back to six pure line ancestors in the USA. Though 77% of a sample of US corn breeders maintained that their base of germplasm was broader in 1981 than in 1970 (Duvick, 1984, Table 16, p. 169), Smith (1988) concluded that there was no change in genetic diversity of Corn Belt maize from 1981 to 1986, and Cox et al. (1988) found that less than 1% of US hybrid corn had non-North American exotic germplasm. Moreover, the National Research Council (1993, p. 73)notes that 'Most surveys have shown that there is little immediate prospect for a large-scale increase in diversity of hybrid maize' in the United States. Apparently, within the germplasm base of US hybrid corn (a small fraction of the total world germplasm), the pool of diversity remains sufficient to provide disease resistance in the high-input US environment and to support a remarkable and as yet undiminished rate of yield increase.
Clearly, genetic resources from the South, made available in recent decades to CIMMYT and other germplasm facilities, have not been of very significant benefit to corn producers in the North. One implication is that the maximum gain to be had by sources of corn germplasm via effective bargaining with Northern corn breeders may be modest indeed, if retroactive compensation is ruled out.
In wheat, for the United States, of 224 wheat cultivars released before 1975, only 31% had any germplasm introduced apart from their foundation germplasm (Cox, 1991, Table 3-1, p. 26), and none of this was introduced later than 1920 (Cox, 1991, p. 28). Of cultivars released subsequently, Cox found that 75% had some more recently introduced parentage, but usually it constituted only a small part of the cultivar's germplasm, typically introduced for disease resistance via crosses and back-crosses. He noted that 'The limited utilization of landraces is most striking ... ' (Cox, 1991, p. 29).
In soybeans, Sprecht and Williams (1984, p. 65) found that of 136 successful soybean cultivars released by US breeders from 1939 to 1981, 121 had cytoplasm from just five introductions. Only six ancestral strains accounted for nearly 60% of the germplasm in these 136 releases and for a similar percentage of germplasm in cultivars released from 1971 to 1981 (Sprecht and Williams, 1984, Table 3-7, p. 68), even though there was large turnover in the set of leading cultivars between 1970 and 1980 (Duvick, 1984, Table 4, p. 164).
In a more recent study, Gizlice etal. (1994, p. 1143) define the genetic base as the 'sets of genotypes that contain 99% of the genes found in modern cultivars'. They conclude that 'the soybean genetic base was largely formed before 1960. Nearly 75% of the genes in modern soybean cultivars is present in sixteen cultivars and a breeding line released before 1960. Breeders have remained dependent on this early genetic core of breeding material and have rarely introduced new germplasm' (p. 1149).
Thus, much of the germplasm of major crops and their wild and weedy relatives already resides in gene banks. But the effect of the vast increase in accessions since the 1970s on germplasm utilized for crop production has thus far been modest.
Allard (1992) offers a breeder's view of the need for an inflow of novel genetic material:
Breeding in barley and corn, as well as in other major crops. has increasingly focused on crosses among elite materials and rates of progress indicate not only that this strategy has been successful but also that there has been little, if any, slowing of progress due to reduction of exploitable genetic material. ... It consequently seems unlikely that readily exploitable genetic variability will soon be exhausted ... (pp. 144-145).
What are the prospects for future crop germplasm demand?
Note that these major crops are precisely the ones with the volume most capable of supporting a competitive private breeding industry. It is sobering that their yields can continue to increase with little introduction of new genetic material into their breeding lines.
A frequent rationale offered by breeders for their low rate of introduction of new genetic material is that cultivars in genebanks or in situ are insufficiently described and documented, so their potential contributions as part of a breeding system can be hard to assess. This is not the whole story, however. Common beans, though a 'minor crop' in the United States, are a staple for millions of people. Moreover their genetic uniformity has led to some disastrous disruptions of production (e.g. the 1982 rust epidemic that caused pinto bean yield losses of 25-50% in Colorado and Wyoming; National Resource Council, 1993, p. 68). Yet
The gap between identification of useful characters in exotic germplasm and the transfer of these potentially useful characters to cultivars had widened. It is economically prohibitive for private companies to commit the time and expense on cultivar development incorporating exotic germplasm in such a minor crop as common beans, and there is no longer much career incentive for public scientists to perform this work. Therefore, the gap ever widens' (Silbernagel and Hannan, 1992, pp. 2-3).
Apparently, the potential prevention of a multimillion dollar disaster offers insufficient incentive for private plant breeders even when well-identified, useful germplasm is available gratis. This gives us some clues as to the extent to which breeders believe they can hope to capture the social value of their work. It also gives us a reality check of the scope of concerns about 'profiteering' by seed companies using germplasm from the 'South'.
Was this article helpful?