| Clearly some reconsideration of the criteria to be used in the establishment
of genomic relationships is in order. Cytologists have always been faced
with the problem of what is sufficient pairing in hybrids to establish homology
on a scale large enough to be considered genomic. The problem is now further
compounded by the recognition of genetic systems influencing the degree
of chromosome pairing in hybrids. In the Triticinae, for example, variation
has been recorded in several species. DOVER and RILEY (1972) have recognized
four classes of pairing in T. tripsacoides (Ae. mutica) hybrids;
DVORAK (1972) and KIMBER and ATHWAL (1972) report three classes in T.
speltoides, while DOVER and RILEY (loc. cit.) record four, but
in the presence of B chromosomes. KIMBER (unpublished) has recovered a "super-high"
pairing type of T. speltoides which differs from the super-high described
by DOVER and RILEY (loc. cit.) in that its presence results in super-high
pairing even in the presence of chromosome 5B. MELLO-SAMPAYO (1972) describes
an intermediate-pairing type in T. longissimum, and recently KIMBER
and SALLEE (1973) have found a high-pairing type in this species also. In attempts to recognize genomic relationships, it would seem that the pairing pattern most likely to give a correct indication of similarity is that in which homoeologous chromosome pairing is reduced to a minimum but homologous chromosome pairing is not affected. Thus the low-pairing classes of DOVER and RILEY (1972) and KIMBER and ATHWAL (1972) and the class I of DVORAK (1972) could be the pairing patterns best indicative of genomic similarity. It is unfortunate that the only type of T. speltoides pairing behavior in hybrids with polyploid wheat recognized prior to 1972 was the high-pairing. Had the accessions of T. speltoides been only of the low-pairing type, it is improbable that it would have been seriously considered as the donor of the B genome of polyploid wheat. In view of these considerations it is perhaps appropriate to present a diagram which may represent our current understanding of the evolutionary pathway of wheat (Fig. 1). This diagram must, in some ways, be considered tentative. It does not, for example, recognize the genomic situation of T. zhukovskii. Similarly, it places the origin of the B genome in some unidentified diploid species. KIMBER and ATHWAL (1972) considered various possibilities for the origin of the B genome. After eliminating such alternatives as an as-yet-undiscovered form of T. speltoides, autopolyploids of the A genome diploids and a hybrid B genome donor, they concluded that the polyphyletic origin of wheat was an alternative that must be considered seriously. Such an origin would most likely have involved two or more independently produced amphiploids involving diploid wheats and unspecified, but diverse diploids. Intercrossing of these amphiploids would cause little change in the A genome component, but might repattern the other genomes so that recognition of the diploid species involved would be impossible. However, there is another possibility: simply that the B genome donor species has become extinct. |
| <-- Back |