General discussion and conclusions

The average total chromosome number in the SB₁ generation was surprisingly low (2n=47). In a backcross to an amphidiploid (AD) T. durum x A. intermedium (2n=70) using A. intermedium (2n=42) as male backcross parent one expects a plant with 2n=56 chromosomes (schematic representation in: SCHULZ-SCHAEFFER 1972). The observed total chromosome number deviates from the expected by 9 chromosomes. In 1958 2n numbers for the AD generation varied from 58 to 74 and averaged 65 (SCHULZ-SCHAEFFER 1970). We assumed that the missing bivalents were from the Agropyron parent since they were foreign to the cytoplasm of the female Triticum parent and may have been eliminated as laggards in meiotic divisions. Since micronuclei had been observed in the AD and our experimental AD material was removed by 7 to 8 generations form the original cross (F₇ and F₈, respectively) such a loss of Agropyron chromosomes may have been possible.

A loss of 9 Agropyron chromosomes without the presence of the 14 T. durum chromosomes would very likely have been fatal. The 14 Triticum chromosomes must have a compensating effect on the AD plant. In hexaploid wheat, T. aestivum L., nullisomics (2n=40) are generally reduced in size, vigor, and fertility. Many can not be propagated. A loss of more than one pair of chromosomes is very likely to be fatal even in a polyploid.

At least two nullisomics in wheat cause partial asynapsis. [Nulli-3A (XII) and nulli-3B (III), SEARS, 1954]. Since synapsis is controlled by genes (BEADLE 1930, 1933, LI et al. 1945, BERGNER et al. 1934, BEASLEY and BROWN 1942, ENNS and LARTER 1960, GOWEN 1933) one would expect that lack of chromosomes can cause partial asynapsis.

The Triticum univalents apparently also have an asynaptic effect on the normal pairing of the Agropyron homologues. Examples from Triticum and Hordeum were given earlier (SCHULZ-SCHAEFFER et al. 1973).

Partial asynapsis is very strongly expressed in the SB₁ generation. The average number of SB₁ bivalents was only 11.7 where 21 were expected. This is also reflected in the observed average number of SB₁ univalents of 21.0 instead of the expected 14.0. This asynaptic effect decreased rapidly in the SB₂ (19.1_II) so that there was little improvement possible in the SB₃ (20.2_II).

Total chromosome number and the asynaptic effect varied from cell to cell within the same plant and from plant to plant in the SB₁ generation. The large variation of chromosome number among the PMC's of the same plant can only be explained as a premeiotic event and may be due to non-disjunction. The high variation in chromosome number from PMC to PMC in turn would cause differences in asynaptic effect which would vary from PMC to PMC. Chromosomes responsible for synapsis may be entirely missing from specific cells causing varying degrees of asynapsis in different cells of the same plant. This facet of meiotic instability was evaluated by the computation of variances of the mean (_Sx²) and pooled variances of the mean (_Sp²/n).

We found that with the progessive stabilization of total chromosome number within plants and between plants from the SB₁ to the SB₃ the bivalents and univalents also stabilized. The final conclusion after evaluation of the SB₂ data (SCHULZ-SCHAEFFER et al. l973) still holds true after evaluation of the SB₃ data. Substitution backcrossing rapidly leads to the normalization and stabilization of meiotic behavior, elimination of Triticum chromosomes, and a more rapid stabilization of bivalents than univalents.

(Received May 28, 1974)