|
The single diploid hybrid reported here is Ae.
longissima Schweinf. et Muschl. x Ae. umbellulata
Zhuk (Table 2). A low level of
pairing was observed with a mean of 2.00 bivalents, 0.36
trivalents, and 0.04 quadrivalents. KIHARA (1945) obtained a
hybrid between these same two species and observed a similar
level of pairing, reporting a range from zero to four
bivalents with a mode of two, none of which were rings, and
no or one trivalent. His conclusion, which the present data
support, was that these two species do not have a common
genome.
Four triploid hybrids were obtained (Table
2). Three had been produced earlier by KIHARA (1937,
1945) and the fourth by KIMBER & ABU BAKAR (1981). For
Ae. ventricosa x Ae. umbellulata, KIHARA
(1937) reported a range from zero to five bivalents, with a
mode of two, and no or one multivalent and concluded that
the two species have no genome in common. The mean pairing
reported here, 2.94 bivalents, 0.22 trivalents, 0.09
quadrivalents, is consistent with this conclusion. Two
haploids of Ae. ventricosa (D and L genomes) were
recently obtained by FEDAK (1983). Very little pairing with
a mean of only 0.46 chiasmata per PMC was reported for the
two plants. However, in the hybrid repored here, there was a
chiasma frequency of 3.66. If as little pairing occurred
between chromosomes of the D and L genomes in this hybrid as
occurred in the haploid, then pairing in the hybrid must be
between the chromosomes of the D and U genomes and the L and
U genomes. This could account for the bivalents but not for
the trivalents, as many as two of which were observed in
some PMCs. A more likely explantation is that pairing was
distributed equally among these three distinct genomes and
that the extent was the result of promotion of heterogenetic
pairing by this genotype of Ae. umbellulata.
For Ae. cylindrica Host x Ae. umbellulata,
KIHARA (1937) reported ranges from six to eight bivalents,
as many as two of which were rings, and zero to three
multivalents. Reviewing this and other related data, KIHARA
(1954) concluded that the U genome of Ae. umbellulata
is a modified version of the C genome of Ae. caudata
L. The pairing in this hybrid was assumed by KIHARA to be
due to pairing between the chromosomes of the U and C
genomes. However, the close relationship between these two
genomes has been questioned by others (SEARS 1948 ;
CHENNAVEERAIAH 1960) with recent analysis suggesting the two
genomes are nonhomologous (KIMBER & ABU BAKAR 1981 ;
MILLER 1981). The hybrid reported here had means of 4.20
bivalents, 1.08 trivalents, 0.23 quadrivalents, and 0.20
larger multivalents (Table 2).
The hybrid Ae. kotschyi Boiss. x Ae.
umbellulata had an average of 5.53 bivalents, nearly
three of which were rings, 0.02 trivalents, 0.58
quadrivalents, and 0.02 quinquevalents. The mean of 0.58
quadrivalents per PMC shows that the two U genomes differ by
a translocation. KIMBER & ABU BAKAR (1981) earlier
reported a hybrid between these two species which had
similar pairing. However, they reported no quadrivalents
which means that a polymorphism for a translocation must
occur either in Ae. umbellulata or more likely in
Ae. kotschyi. FURUTA (1981) has already shown the
presence of such a polymorphism in his collections of Ae.
kotschyi.
The karyotype of Ae. umbellulata has four markedly
heterobrachial chromosomes (SENJANINOVA-KORCZAGINA 1932 ;
CHENNAVEERAIAH 1960) that can be readily identified if they
are involved in a multivalent configuration at MI. A
heterobrachial chromosome was involved in only three of the
thirty-three observed quadrivalents in Ae. kotschyi x
Ae. umbellulata. Thus, the translocation does not
involve any of the heterobrachial chromosomes of the U
genome of Ae. kotschyi.
The fourth triploid hybrid, Ae. triuncialis L. x
Ae. caudata, has an average chiasma frequency similar
to that of the previous hybrid and a mean of 6.11 bivalents,
1.00 trivalent, and 0.07 quadrivalents (Table
2). KIHARA (1945) reported for this hybrid combination a
mode of seven bivalents, as many as six of which were rings,
and zero to two multivalents. The pairing in a haploid
reported by CHAPMAN & MILLER (1977), most likely derived
from the same accession of Ae. triuncialis as that
used here, indicates that at least two bivalents as well as
a trivalent or quadrivalent in this hybrid may be accounted
for by pairing between chromosomes of the U and C genomes.
Even though KIHARA (1954) maintained that the C genome was
present in an unmodified condition in Ae.
triuncialis, there is evidence that modification has
occurred. Both SENJANINOVA-KORCZAGINA (1932) and
CHENNAVEERAIAH (1960) have shown that the karytype of the
presumed C genome in Ae. triuncialis is altered in
some varieties. The pairing reported here for this triploid
hybrid suggests that such modification has occurred in this
accession of Ae. triuncialis.
|