Genetic compatibility among domesticated and wild accessions of the tropical tuberous legume Vigna vexillata (L.) A. Rich.

Abstract

Phenotypic similarities and differences between a cultivated variety of the tuberous legume Vigna vexillata from Bali, Indonesia, the putative domesticated variety macrosperma and wild types from Africa, Australia and Indonesia, were reported previously. The present study was undertaken to establish the genetic compatibility of these three accession classes. Seventeen accessions, comprising eight cultivated Bali accessions, one var. macrosperma accession and eight wild accessions from Africa and Austronesia, were grown in large pots in shade house facilities in Townsville, Australia. Not all hybrid combinations were attempted because for some accession combinations, suitable matching flowers were not available at the same time. The main aim was to attempt enough crosses between accessions from the respective classes to establish whether the classes were genetically compatible. Hybridisation was conducted by hand pollination in the morning, using newly-open flowers that had been emasculated before sunset on the day before. Pods and viable hybrid seed were obtained from the Bali × Bali, var. macrosperma × wild and wild African × wild Austronesian combinations. However, difficulty was encountered in obtaining viable and/or self-fertile hybrids between the Bali accessions and the other two classes. Depending on the particular combination of parental accessions, different genetic breakdown mechanisms were observed with the Bali × var. macrosperma and Bali × wild combinations. In some instances, flowers failed to set pods and/or the young pods abscised before maturity; pods set but seed were shrivelled and/or non-viable; viable seeds were set but the hybrid seedling plants were short-lived; or, in a few instances (Jimbaran Bali × wild Austronesian), vigorous hybrid plants were obtained but were self-sterile. Mitotic chromosome counts showed there was no difference in chromosome number between the Bali accessions, the Austronesian accessions and those hybrids that were viable but infertile. All exhibited 2n = 22. Pollen viability analyses using Alexander’s stain indicated that the numbers of pollen grains per flower and the percentages of pollen grain that were viable were substantially lower in the hybrids than in both the cultivated Bali and wild parental accessions. Consistent with this observation, small numbers of viable seeds were obtained when viable pollen from the respective parents was backcrossed onto the self-sterile hybrids. The results suggested that the cultivated Bali accessions do not belong to the same primary gene pool as the other cultivated and wild V. vexillata accessions and that it would be difficult to use the Bali accessions and var. macrosperma concurrently to breed seed crop varieties. A taxonomic review of V. vexillata is also warranted.