HYBRIDIZATION IN THE GENUS GLYCINE SUBGENUS GLYCINE WILLD. (LEGUMINOSAE,PAPILIONOIDEAE)

The genus Glycine is composed of two subgenera, Glycine and Soja. Soja includes the cultivated soybean, G. max, and its wild annual counterpart G. soja, while Glycine includes seven wild perennial species. Hybridization was carried out within and between wild perennial species of the subgenus Glycine. The success rate (pods set/flowers crossed) was 11% for intraspecific and 8% for interspecific crosses. A total of 220 F₁ hybrids was examined morphologically and cytologically where possible. Hybrids within G. canescens (2n = 40) and G. latifolia (2n = 40) were fertile as expected. Glycine clandestina (2n = 40) was morphologically separable into at least three groups, which produced fertile hybrids within each group. One cross between two groups gave vegetatively vigorous but sterile hybrids. The majority of crosses within G. tabacina (2n = 80) were fertile, except that extremely narrow-leaved forms gave sterile hybrids in combination with more usual forms. Sterility was also encountered in G. tomentella when aneuploids (2n = 78) from New South Wales, Australia, were crossed with tetraploids (2n = 80) from either Queensland, Australia, or Taiwan; crosses between the latter two populations resulted in seedling lethality. Cytological behavior of sterile hybrids followed a similar pattern, whether at the diploid or tetraploid level. The frequency of chromosome pairing was approximately half that expected if genomes showed full pairing homology. Bivalent disjunction at anaphase I was usually followed by precocious division of the majority of univalents. Telophase I and II were characterized by lagging chromosomes and micronuclei, so that resulting pollen was misshapen and sterile. Chromosome pairing data from sterile intraspecific hybrids at the tetraploid level may indicate a polyphyletic origin of tetraploids, whereby different diploid populations were involved in their formation. Similarly, chromosome pairing in sterile intraspecific diploid hybrids may indicate that the various diploid groups arose independently of one another. Both 40- and 80-chromosome forms are fully diploidized, however, and if they are of ancient origin, divergence since that time could have resulted in the chromosomal differentiation which becomes apparent when intraspecific hybridization is effected. Diploid (2n = 40) interspecific hybrids G. falcata × G. canescens, and G. falcata × G. tomentella grew poorly and did not reach flowering stage. Diploid (2n = 40) crosses between G. latifolia and G. tomentella produced inviable seedlings. Tetraploid (2n = 80) hybrids between G. tomentella and G. tabacina were vegetatively vigorous but sterile owing to low chromosome pairing at meiosis, indicating little pairing homology between the two species. Diploid hybrids between G. canescens and G. clandestina, however, showed almost complete chromosome pairing at diakinesis and partial fertility. Although morphologically distinct, these two species have not diverged sufficiently to prevent hybridization and possible gene exchange through recombination. Self compatibility, perennial growth habit, and geographic isolation have favored divergence among Glycine populations to the point that gene exchange appears no longer possible in many cases. Internal isolating mechanisms have been shown to operate at various levels of plant development from hybrid lethality at seedling stage, to failure of seed-set in sterile but vegetatively vigorous hybrids.