Crosses between barley (Hordeum vulgare)
and bulbous barleygrass ( H.bulbosum)
could be valuable for the transfer of such properties as resistance to cold or diseases from H. bulbosum
to H. vulgare.
From the literature it was known that difficulties arose in the cross: seed abortion necessitating culture of the hybrid embryo, sterility of the hybrids, and all the viable hybrids mentioned in the literature were triploid. The purpose of my study was to enlarge knowledge of these interspecific crosses, and to try and solve the difficulties. For convenience parents and hybrids were denoted by genome symbols, e.g. VV = H. vulgare
(2x), BB = H. bulbosum
The results below are arranged as in the chapters.Crossing and the hybrids
All eight possible crosses between diploid and autotetraploid cytotypes of the two species were tried. The direction of the cross, the parents and environmental conditions seemed to influence its success. Both crosses between VV and BB yielded three haploid plants with the genome V and six diploid hybrids. The haploids and two of the hybrids died before flowering. The cross VV x BBBB was most successful: one haploid plant with the genome V and 145 triploid hybrids, of which 13 died before flowering. The reciprocal cross yielded 12 triploid hybrids. Both crosses between VVVV and BB were unsuccessful, whereas the two crosses between VVVV and BBBB produced 18 dihaploids with the genome VV and 16 plants with more than 14 chromosomes and hybrid characteristics. Of the dihaploids nine died before flowering and of the hybrids six (tables 2 to 6).
All the hybrids resembled one another and were almost sterile (tables 7 and 8). Their general form was that of H. bulbosum,
although they were not so large and vigorous. Chromosome number was usually unstable in the tetraploid hybrids; from one of them a diploid hybrid arose, presumably by gradual elimination of chromosomes, while six of the plants formed tillers, usually fertile, resembling H. vulgare.
The progeny of these tillers were diploid with the exception of one which yielded diploid and tetraploid plants, presumably by spontaneous doubling up of the chromosomes.
The dihaploid plants closely resembled diploid H. vulgare
and were fertile. They and the haploid must have arisen by elimination of chromosomes. Later generations from dihaploids and from fertile tillers of tetraploid hybrids were compared with H. vulgare.
Progeny with cytoplasm of H. vulgare
(from the cross VVVV x BBBB) differed slightly from diploid H. vulgare,
probably because the genetic constitution of the two sets of chromosomes derived from the tetraploid is no longer identical with that of the diploid variety of barley. Offspring of dihaploids with the cytoplasm of H. bulbosum
(from the cross BBBB x VVVV differed more because there would be a cytoplasmic effect as well as the genetic effect (table 9).Embryology
Development of the caryopsis from the crosses VV x BB, VV x BB13B and VVVV x BBBB was compared with intravarietal fertilization VV x VV. All parts of the caryopsis developed slower in interspecific crosses. Seed of the crosses VV x BB and VVVV x BBBB usually aborted within two weeks of pollination before any organs of the embryo had differentiated. Seed of VV x BBBB aborted within three or four weeks after organs of the embryo had started differentiating abnormally.
From early in development all interspecific crosses showed abnormalities in division of cells and nuclei in the endosperm. Besides many micronuclei, there were multinuclear cells and giant nuclei which arose by endomitosis or by nuclear fusion in the multinuclear cells followed by endomitosis. The abnormal endosperm degenerated, thus appearing to be the primary cause of seed abortion. In the hybrid embryo abnormalities were detected, from which micronuclei arose. They appeared later than in the endosperm and seemed not to be the direct cause of abortion of the embryo, even though they affect its vitality.
In the hybrid crosses the antipodal cells remained intact slightly longer than in the intravarietal cross and occasionally some cells remained intact much longer, giving rise to enormous giant nuclei, perhaps by endomitosis.
The maternal tissue in the hybrid crosses developed initially as in the intravarietal cross but more slowly. Seed abortion certainly curtailed this early.Cytogenetics
The karyotypes of parental species, hybrids and various other progeny were described and compared, to see whether there was any change in chromosomal morphology and to find the chromosomal composition of plants after vegetative elimination of chromosomes. Amphiplasty occurred in the hybrid so that satellites of the H. bulbosum
genome were not visible as such. The relative lengths of the chromosomes within a genome were not affected by hybridization and the ratio of lengths between the genomes of H. vulgare
and H. bulbosum
were about the same in all hybrids. The karyotypes obtained by vegetative elimination of chromosomes were entirely as expected (tables 12 to 22).
The maximum number of nucleoli per cell of vegetative tissue corresponded with the number of satellite chromosomes except in some cells of triploid hybrids. The average number of nucleoli per cell was in diploids and triploids much larger and
in tetraploids slightly larger than expected from the number in H. vulgare
. Thus the presence of chromosomes of H. bulbosum
increased the number of nucleoli per cel (table 23 and figure 26).
In all types of hybrid, chromosomes paired completely or almost completely during the pachytene stage of meiosis, by which allosyndetic, non-homologous autosyndetic and homologous autosyndetic pairing occurred simultaneously. Then some associations of chromosomes fell apart by desynapsis which was a result of low chiasma frequency. The other chromosomes remained joined by chiasmata or pseudochiasmata, which could hardly be distinguished from each other. In diploid hybrids most seemed to be pseudochiasmata and they sometimes lasted until the first anaphase. In the triploids most normal chiasmata gave rise to association of chromosomes in the first metaphase, while in tetraploids both types occurred together but with fewer pseudochiasmata than normal chiasmata (tables 24, 25, 28, 29, and 31).
In all types of hybrid there were one or rarely two nucleoli per cell during prophase and diakinesis of meiosis. The nucleolus was associated with two chromosomes or chromosomal configurations at most.
In diploids and tetraploids and to a lesser extent in triploids there were aneuploid cells, chiefly hypoploid. Many cells contained various inclusions; hypoploid cells contained the most. The inclusions and the aneuploidy could be ascribed to abnormal premeiotic division in which not all the chromosomes were distributed between the daughter cells (tables 26 and 32).
Association of chromosomes was studied in the first metaphase of aneuploid cells. In diploid hybrids, association decreased rapidly in hypoploid cells with declining chromosome number. In triploid hybrids this decrease was less marked and in cells with about 14 chromosomes association reached the rate found in the diploid hybrids, suggesting that it was chiefly the chromosomes of H. bulbosum
which had been eliminated. In tetraploid hybrids association decreased rapidly until chromosome number reached 24 and then decreased slower slightly above the rate of association found in aneuploid cells of triploids. This indicates that there is some selectivity in the elimination of chromosomes, whereby the chromosomes of H. bulbosum
stand more chance of being eliminated (figure 33).
Because of univalents, pseudochiasmata and aneuploidy, meiosis was very irregular from the first metaphase. The tetrads contained micronuclei and pollen micrograins. Ripe anthers of diploid hybrids contained no stainable grains; triploids contained little stainable grains and they were of variable diameter; in tetraploids there was a proportion of stainable grains which varied between plants and grain diameter was very variable (tables 27, 30 and 34).
In the haploid plant with the genome V there was a slight tendency to form bivalents in meiosis, but they mostly proved to be pseudobivalents. Dihaploids with the genome VV all had very regular meiosis similar to that of diploid H. vulgare
. The pollen of these plants was identical with that of diploid H. vul
gare (tables 33 and 34).Discussion
Many of the abnormalities seem to be caused by faulty division of nuclei containing chromosomes from both H. vulgare
and H. bulbosum.
They included the abnormal development of endosperm, lack of vigour in the hybrid embryo and plant, vegetative elimination of chromosomes, and- the occurrence of aneuploid pollen mother cells. The abnormalities could be the result of incompatibility between chromosomes or genes of the parental species. The observed differences in the amount and nature of abnormalities in several tissues probably ensue from existing differences between the tissues, e.g. in speed of mitoses, and by differences in gene dosage.
Another abnormality in the hybrid nucleus is amphiplasty, which was expressed as the suppression of the secondary constriction in the satellite chromosome of H. bulbosum
and probably caused the suppression of nucleolus activity in the same chromosome. This abnormality is perhaps associated with the preferential elimination of the chromosomes of H. bulbosum.
Finally there are abnormalities in meiosis: besides aneuploidy, there are desynapsis as a result of low chiasma frequency and formation of pseudochiasmata. The low chiasma frequency probably is genetically determined, while the pseudochiasmata might be the result of matrix connections or fusion of heterochromatin.
The significance of the hybrids for barley breeding is briefly considered. As it is likely that in meiosis material is exchanged between chromosomes of the two parents and as the hybrids are not completely sterile, it may be possible to introduce characteristics of H. bulbosum
into H. vulgare.
In somatic tissue characteristics could be transferred by induction of interspecific translocations, for example in tetraploid hybrids, after which the preferential elimination of chromosomes could be used to recover changed genomes of H. vulgare.
Perhaps hexaploid hybrids VVBBBB could be used. Plants with the cytoplasm of H. bulbosum
and the chromosomes of H. vulgare
could be of interest for breeding frost-resistant barley and in the study of male-sterile barley. The possibilities deserve further research.