been confused with the areas of evolution and diversity of crop plants (Hawkes, 1983). Harlan (1951) identified smaller microcenters that are rich in diversity and not necessarily near civilization nor confined to plains or mountains—features strongly associated with the Vavilovian centers—usually in the areas that would be regarded as broader centers of diversity.
There are therefore different definitions of the regions that comprise centers—from Vavilov’s to those of Zhukovsky (1975). Whatever the definitions, it is still clear that there are areas where crop plant diversity is great and others where it is slight. The centers of diversity of several crops coincide and comprise limited numbers of regions in contrast to the 13 centers proposed by Vavilov. Crop germplasm is by no means obtained solely from developing countries; centers also cover areas of the developed world. Examples of crops that originated in the developed world include the adzuki bean (Vigna angularis) in Japan, oats (Avena sativa) and rye (Secale cereale) in northern Europe, and the sunflower (Helianthus annuus) in the United States. Many old landraces of the developed world are extremely valuable resources being fed back, through breeding, to Third-World agriculture.
In view of the apparent confusion surrounding the definition of the regions of diversity, it is pertinent to question how the material can best be identified and preserved. Over the past decade IBPGR’s strategy has been to mobilize scientific information from hundreds of the world’s best breeders and crop botanists. Scientists who have grown, used, and evaluated germplasm know far more about its variability than collectors who can only observe phenotypes. This strategy has been highly successful in ensuring the preservation of intraspecific diversity in a very limited time and is still implemented.
For the wider gene pools, where species and species relationships are the principal interest, taxonomy plays a major role in the conservation of genetic resources work, especially experimental taxonomy designed among other things to clarify genomic relationships. But information on the origins and evolution of crop gene pools is obtained not only from taxonomy but also from a synthesis of disciplines, including archaeology, linguistics, ecology, and molecular biology.
For some of the major crops, e.g., wheat and maize, there are still gaps in our knowledge, but in many other cases, the progenitors are well known. Although the ancestor of the faba bean (Vicia faba) has not been equivocally identified, several closely related species have frequently been found, the most recent during an IBPGR mission to Syria in 1986. Truly wild forms of cassava are not definitely known either, although Brazilian workers supported by IBPGR may have found one in 1986.
When a theory of origin has been postulated but not verified by taxonomic work, it may not be clear what germplasm should be collected and where. The cowpea (Vigna unguiculata) is a good example; taxonomically its ancestral home has been variously proposed as Ethiopia or central/southern Africa. IBPGR examined all herbarium specimens of this species in 1986 and found that the progenitor of the cowpea occurs in every country of sub-Saharan Africa. Thus until experimental work on freshly collected material is complete, we will not know the patterns of variation or where to collect samples. Meanwhile, collection of material