|
The single tetraploid hybrid obtained was between Ae.
triuncialis and Ae. cylindrica. Observed were an
average of 7.29 bivalents, of which only 2.25 were rings,
1.16 trivalents, 0.39 quadrivalents, and 0.15 larger
multivalents (Table 2). Higher
bivalent frequencies were reported in earlier hybrids
between these two species. PERCIVAL (1930) recorded a range
from three to eleven bivalents of which five to six were
rings. KIHARA & LILIENFELD (1932) reported zero to
twelve bivalents with a mode of ten to eleven ; three to six
were rings. KAGAWA (1931, cited in KIHARA 1937) in the
reciprocal combination reported six to eleven bivalents and
no or one trivalent. The possible modification in the C
genome of this accession of Ae. triuncialis discussed
above could account for the apparent discrepancy between the
pairing reported here for this tetraploid hybrid and that
from earlier reports.
In the three pentaploid hybrid combinations, Ae.
recta (Zhuk.) Chenn. was a common parent. The second
parent in each case possessed the U genome in common with
Ae. recta. Observed in Ae. variabilis Eig x
Ae. recta were means per PMC of 6.63 bivalents, 2.27
of which were rings, 1.19 trivalents, and 0.11 quadrivalents
(Table 3). KIHARA (1945) reported
seven to eleven bivalents, as many as four of which were
rings, and zero to two multivalents. He concluded that the
two species had only one genome (U) in common. The
chromosomes of the B and U genomes were shown to be able to
produce at least two bivalents in the diploid hybrid
discussed above. This affinity may be reflected in this
pentaploid by the frequencies of both rod bivalents and
trivalents.
In addition to the U genome, the other pentaploid hybrid
combinations supposedly have versions of the M genome in
common. Evidence for this is indirect and in no case has an
unmodified M genome from Ae. comosa been demonstrated
in an Aegilops polyploid. In the case of Ae.
geniculata Roth (=Ae. ovata auct.) evidence for a
modified M genome is very indirect. In fact, the extent of
pairing in hybrids of Ae. geniculata and Ae.
ventricosa (e.g., KIHARA 1937) suggests that these two
species have a genome in common, in which case the genome
formula of Ae. geniculata should be LU. There is even
less direct evidence for an M genome in Ae. lorentii
Hoch. (=Ae. biuncialis Vis.). Based on karytype
analysis CHENNAVEERAIAH (1960) concluded that the second
genome of Ae. lorentii might well be something other
than a modified M.
Aegilops recta x Ae. geniculata hybrids have
not been reported before. The three plants reported here had
means of 8.88 bivalents, 2.76 of which were rings, 1.10
trivalents, 0.21 open quadrivalents, and 0.02 quinquevalents
(Table 3). In the Ae. recta
x Ae. lorentii hybrids, means of 7.94 bivalents, 2.04
of which were rings, 1.79 trivalents, 0.24 open
quadrivalents, 0.15 quinquevalents, and 0.01 sexivalents
were observed (Table 3). KIHARA
(1945) reported two hybrids from this parental combination.
One had about nine bivalents, as many as five of which were
rings, and some trivalents and quadrivalents. The other had
a range of six to eleven bivalents, many of which were
rings, and an occasional quinquevalent. These results for
these two hybrid combinations provide evidence for no more
than one genome in common in each case as was found for the
Ae. variabilis x Ae. recta combination just
discussed. In conjunction with the karyotypic evidence of
CHENNAVEERAIAH (1960), this must be the U genome. Thus, the
identity of the non-U genomes in these polyploids is
questionable.
The hexaploid hybrid, Ae. juvenalis (Thell.) Eig x
Ae. recta, is also a combination reported for the
first time. In the two plants there were combined means per
PMC of 7.25 bivalents, with 1.17 rings, 2.10 trivalents, and
0.29 quadrivalents. On the basis of these data, especially
the low frequency of ring bivalents, one might question if
one genome could be considered common to both species.
However, other evidence has shown the presence of the U
genome in both Ae. juvenalis (MCGINNIS 1956 ; KIHARA
1963 ; KIMBER & ABU 1981) and Ae. recta (KIHARA
1937, 1957).
None of the data suggested any definitive changes in the
current understanding of the identity of the component
genomes of the polyploid taxa considered here. However,
there is enough uncertainty in the cases of the polyploid
species that presumably have modified M genomes to warrant
some direct genome analytical work particularly with the
diploid species Ae. comosa and Ae. uniaristata
as testers as was done recently with Ae. uniaristata
and Ae. ventricosa (KIMBER et al. 1983). In
the cases here where comparisons could be made with earlier
hybrids, the discrepancies underscored the need for wider
sampling within species to assess the extent of genome
modification and of genetic variability for the requlation
of heterogenetic pairing. In addition, the level of pairing
in the Ae. ventricosa x Ae. umbellulata hybrid
plant when compared with pairing reported in haploids of
Ae. ventricosa (FEDAK 1983) suggested promotion of
heterogenetic pairing by Ae. umbellulata.
Acknowledgements
The author wishes to acknowledge review and discussion of
the data by JAN DVORAK and GORDON KIMBER.
|