CHROMOSOME BEHAVIOUR IN F1 WHEAT HYBRIDS: I. PENTAPLOIDS

Meiosis was studied in varieties of Triticum vulgare (2n = 42), T. dicoccum (2n = 28), T. durum (2n = 28), T. Timopheevi (2n = 28), and in 16 of their pentaploid hybrids as part of a study in an attempt to establish criteria indicating relationships between 42- and 28-chromosome wheats, with particular reference to the possible relationship of the new 42-chromosome wheat, McMurachy's Selection, to T. dicoccum or T. durum.One plant each of T. vulgare var. Hope and Marquillo had only 41 chromosomes. One plant of T. durum var. Pentad had three times as many unpaired chromosomes as the other plants of this variety.A nucleus with 14 pairs and 7 univalents was not detected among the 86 pollen mother cells analysed in the cross involving T. Timopheevi. In the remaining crosses the frequency of this association of chromosomes was lowest in the three hybrids involving T. durum var. Pentad, greater in the three involving T. dicoccum var. Khapli, still greater in the three involving T. dicoccum var. Vernal, and greatest in the nine hybrids involving T. durum var. Iumillo.Of the seven "extra chromosomes" of T. vulgare only six remained unpaired in some pollen mother cells of the hybrids involving Vernal or Iumillo and five in those involving Khapli or Pentad. One pollen mother cell of F1 Marquis × Pentad contained only four unpaired chromosomes.Associations of four chromosomes were rare in some, and not seen at all in others, of the hybrids involving Vernal or Iumillo, more frequent in hybrids involving Khapli, and very frequent in hybrids involving Pentad. In the latter, from 47 to 57% of the nuclei had from one to three such multiple associations, and even chains of five and six chromosomes were observed.Fragmentation of unpaired chromosomes at or in the spindle fibre attachment region was observed in a number of first anaphase figures.There were statistically significant differences in the frequencies of occurrence of micronuclei in tetrads of the 15 hybrids studied at the second reduction division.The crosses R.L. 1544 (genetically related to T. durum var. Iumillo) × Iumillo and Hope (genetically related to T. dicoccum var. Vernal) × Vernal were used as standards for comparison. On the basis of the results, the following criteria were used in attempting to establish relationships between the other 42- and 28-chromosome wheats: (1) the percentage of pollen mother cells with 14 pairs and 7 univalents (greatest in the hybrids between related varieties); (2) the average number of chromosomes involved in multiple associations (lowest in hybrids between related varieties); (3) fertility (greatest in hybrids between related varieties). McMurachy's Selection appeared to be most closely related to T. durum var. Iumillo. On the basis of Criteria (1) and (2), Marquis appears to be more closely related to T. dicoccum var. Vernal than to T. durum var. Iumillo, but in respect of fertility it seems closer to the latter.Chromosome behaviour in the 16 hybrids cannot be neatly summarized. Even varieties within a species gave different results—results that are not in agreement with earlier published reports on chromosome behaviour in pentaploid wheat hybrids in which it has been stated that 14 bivalents and 7 univalents are most commonly found. The difficulties encountered in attempting to establish criteria indicating relationships between the 42- and 28-chromosome wheats suggest that the utmost caution must be used in drawing phylogenetic conclusions on the basis of such data.