2. From wheat cytogenetics to Agropyron study
Genetic study in Japan started when genome analysis of Triticum was extended to the closely related genera and then Agropyron (see Nishiyama 1954). That was by means of cytological observation and had been carrying out during the period of 1920's to 1940's in Kyoto. As stated previously, since the classification of some indigenous species had not been clear for long, necessity was strongly felt for cooperative study between both field of specialities, cytogenetics and grass classification. This incipient motivation to clear up the indigenous species lasted for many decades up to the present days.It is clear from Table 1, classification
had complications in A. tsukushiense and, also A. ciliare.
Consequently, at least, it had been noticed that the differences of characteristics
which were observed among strains of the A. ciliare group and those
of the tsukushiense group, are just of the same degree of difference
as seen among the cultivars of wheats and barleys.
3. Early study of cross hybridization and cytological s.observations
Chromosome studies by 1940 found both hexaploid (6x) and tetraploid (4x) species (Nakajima 1936 and see Matumura S (1941). Matsumura S (1941) carried out crossing experiments within the genus Agropyron and with Triticum using Agropyron species, either Japanese, or introduced species, which had been received from Canada, A. elongatum from K. W. Neatby, and A. glaucum from J. M. Armstrong during the later 1930's (Matsumura S 1941, 1948). It was found that hybridization was possible within Japanese indigenous species and also within the section Elytrigia and equally between the species of section Elytrigia and Triticum, and was unsuccessful when Japanese species were tried with the Elytrigia species or with Triticum.The F1 hybrid between two Japanese Agropyron species, A. ciliare x A. tsukushiense var. transiens (A. semicostatum in Matsumura S 1941, 1948), becomes pentaploid (2n=5x=35). The amount of chromosome pairing showed the presence of two common genomes between the parents, just as between the hexaploid and tetraploid wheats. However, different from wheat, the Agropyron pentaploid hybrid was completely sterile. Genetic reason for the sterility has not been studied since Matsumura S (1948) first noticed it. In his observation of chromosome pairing in the other two cross combinations regarding A. tsukushiense as one parent to A. humidorum (2n=6x=42) and A. cristatum (2n=2x=14) (Matsumura S 1942, 1948), no good MI cell was obtained in the former hybrid. From the hybrid plant (2n=4x=28) of the latter combination, the amount of bivalent formation was low, ranged from 28' to 1'''+ 3'''+ 18', indicating no common genome present between the parents.
In the cross combinations invos.lving Agropyron intermedium (A. glaucum, 2n = 6x = 42, introduced from the Voroney* Botanical Garden, USSR) with tetraploid wheats, the amount of chromosome pairing compared between the F1 plants with different genome constitutions, one involved the B genome of wheat (2n = 4x = 28, AABB, Triticum turgidum), and the other G genome(2n=4x= 28, AAGG, T. timopheevi). Matsumura S (1949) found that pairing was higher in the former hybrids than that in the latter. An average of 7 bivalents in the former hybrid was assumed to be between B genome chromosomes of the parents. However, that assumption was not valid, because in the F1 between an autotetraploid line of T. aegilopoides (2n=4x=28, AAAA) as female to A. intermedium, the extent of chromosome pairing was higher than at autosyndesis of two homologous A genomes from the female parent (Muramatsu 1955a, b). Besides, also the formation of bivalents, and their variations among cells indicated strongly the existences of genetic system controlling chromosome pairing. Bivalents appeared in the F1 Aegilops cylindrica (CCDD) x A. intermedium and Ae. variabilis (UUSlSl) x A. intermedium with chromosome paring of 5''+ 25'∼ 10''+15', and in F1, Ae. longissima (SlSl) x A. intermedium chromosome parings were 1'''+5''+15'∼ 10''+ 8' (Matsumura S and Muramatsu 1956; Muramatsu and Sakamoto 1956; Matsumura S, Muramatsu and Sakamoto 1958a, b). Since from these hybrids the B genome is absent, the involvement of B genome is excluded in Agropyron (Elytrigia group). Then we gave A. intermedium a genome symbol EEFFJJ.
Inconsistently, however, the highest number of the associated chromosomes in the F1, T. durum 8x (2n=56, AAAABBBB) and A. trichophorum (=intermedium,2n=6x=42), was lower than those observed in the hybrids of the cross combinations involving Aegilops(Muramatsu 1956). Those results indicated the involvement of the genotype(s) to control the chromosome pairing. Talking with Dr. Okamoto, when I was studying under the late Dr. E. R. Sears on the monosomic series of wheat, the presence of a genetic system was taken into consideration, and together with his result of asynaptic effect of 5B (Okamoth 1957), the idea was supported.