6. Studies on cross compatibility using Japanese species and cytological observations -chromosome configurations, amphiploidy, polyhaploidy, and anomalous cell-divisions
(1) Agropyron - Triticum hybrid by embryo rescue
Production of viable hybrids involving Japanese Agropyron (Roegneria) species with wheats (Triticum)was realized by means of embryo rescue. Usually seeds formed by pollinating Roegneria species as female with wheat pollen, cease to develop in about a week to ten days, at most 2 weeks after pollination, and then become abortive, shriveled, empty and are unable to germinate.

Takahashi and Muramatsu (1981) by pollinating 2,203 flowers of A. tsukushiense as female by pollen of Triticum aestivum cv. Chinese Spring obtained 214 (9.7%) seeds. Two F₁ hybrid plants were developed out of 100 embryos cultured, while 70 seeds without embryo culture did not germinate. Hybrids (2n=6x=42; ABDSHY) showed good growth and were morphologically intermediate between the parents. Chromosome pairing was compared between tillers divided from one plant and grew under different conditions. Up to 5 bivalents were observed and 49% of the cells had 42 univalents in tillers kept under favorable condition, 20°C and 12 hrs of light, while under the unfavorable varied conditions of 2°C ∼ 25°C and 14∼ 24 hrs of light, 95%cells showed 42 univalents. Although there was very low homology between the 6 genomes involved, apparently chromosome pairing is promoted under the former conditions and one ring bivalent was observed, while the rest were rod-shaped weakly paired bivalents.

Similarly, the hybridization with wheat was possible using A. ciliare as female. Muramatsu and Kaneta (1982) obtained hybrids, using cultivars of T. aestivum, cv. Inayamakomugi and cv. Chikuzen. With A. tsukushiense as female, a hybrid plant was obtained by pollination with T. aestivum cv. Gabo besides with cv. Chinese Spring. Chromosome pairing of F₁ involving A. ciliare showed similar tendencies that were observed in the hybrid involving A. tsukushiense in the previous year, but one trivalent association was seen in the combination with A. ciliare, although its frequency was very low.
(2) Agropyron - Hordeum hybrid by embryo rescue, chromosome number instability and colchicine treatment
Embryo rescue was used for producing hybrid between Agropyron and Hordeum bulbosum (2n=4x=28) (Shigenobu and Sakamoto 1981; Shigenobu-Kishimoto and Sakamoto 1982). In their cross combination, A. repens (2n=6x=42) and A. tsukushiense as female were pollinated with H. bulbosum (4x). A peculiarity appeared in the hybrid plants as there was variation of somatic chromosome numbers. In A. repens x H. bulbosum, somatic chromosome number in root tips was 2n=34 instead of the expected 2n=35, and spike formation delayed to the fourth year after seedling stage, despite prolific tillering heading was poor and in the root tips 1 or 2 chromosomes were reduced. Chromosome numbers counted at MI showed a range of variations from 34 to 30, and bivalents, which displayed autosyndesis of chromosomes of each parent, varied from 7 to 1. In the hybrid, A. tsukushiense x H. bulbosum (4x), in addition to the expected 2n=5x=35, PMCs with reduced chromosome number of 2n = 32 were found and bivalents varied from 8 to 2, indicating that almost all of the pairing was by autosyndesis of the H. bulbosum chromosomes. However, 8 bivalents per cell and appearance of trivalents and quadrivalent were observed, indicating that some chromosomes from both parents might have contributed to the pairing.

They reported another cross A. elongatum (2x) x H .bulbosum (4x). One plant obtained had a morphology similar to that of the female parent, but had 2n=15 chromosomes. Apparently H. bulbosum chromosomes were eliminated and formation of trivalent indicated trisomic nature of the resulting plant (Sigenobu-Kishimoto and Sakamoto 1984).

Chromosome instability is frequently found in certain cross combinations and/or in the backcross generation plants of F₁, A. ciliare x hexaploid wheat (2n = 35), as female pollinated by wheat (2n=42). Kaneta and Muramatsu (1983a, b) reported wide variation of chromosome numbers among PMCs. This was certainly not artifact in making microscopic preparations.

Treating tillers of F₁ plants with colchicine solution, fertile sectors appeared and set some seeds, and of the five amphiploid plants obtained three had the expected chromosome number (2n = 70). From the remaining two, one was 2n=69, and another 2n=66 and in this generation chromosome variations were not observed in root tip cells (Kaneta and Muramatsu 1983a).