Diploidization and chromosomal pairing affinities in the tetraploid wheats and their putative amphiploid progenitor.

The genomes of the diploid wheats Triticum boeoticum and T. urartu are closely related, giving 7II in the f1 hybrid (T(b)T(u)) and 8.4 (0-14) II + 2.5 (0-7) IV in the derived amphiploid (T(b)T(b)T(u)T(u)). The genomes of the tetraploid wheats are also closely related, giving up to 7II at the polyhaploid level (AB) in the absence of the gene Ph but 14II at the tetraploid level (AABB) in the normal presence of Ph. If the amphiploid is the progenitor of the tetraploids, one or the other homoeologue (T(b) or T(u)) in each of the 7 homoeologous groups (the 7 potential IV) must have differentiated with respect to pairing affinity in order to account for 14II in the tetraploid. Consequently, in tetraploid X amphiploid hybrids (T(b)T(u)AB) carrying the Ph gene from the tetraploid, the seven differentiated chromosomes (B) would be expected to give 7I while, on the basis of their observed chiasma frequency, T(b), T(u) and the less differentiated A would be expected to give 4.17I + 3.57II + 3.23III), assuming homoeologous pairing. The expected chromosomal configuration freqencies at MI (11.17I + 3.57II + 3.23III) closely fit the observed values (11.22I + 3.45II + 3.19III + 0.071IV) for such hybrids (X(2) = 0.0046; P>0.99). Thus diploidization of the boeoticum-urartu amphiploid clearly could account for the origin of the tetraploid wheats. Furthermore, T. aestivum X amphiploid hybrids (T(b)T(u)ABD) with and without Ph indicated that B as well as A chomosomes tended to pair with their presumed T(b)T(u) homologues in the absence of Ph. Other tests showed that the tetraploid wheats could not plausibly have originated from any postulated Triticum-Sitopsis (TTSS) parental combinations with or without such chromosomal differentiation.