Recognition of a phenotypically distinct ‘French-type’ plantain (Musa AAB) designated ‘Superplatano’ (Superplantain) prompted evaluation of in vitro micropropagation as a means of generating sufficient numbers of plants for field evaluation in three locations in Puerto Rico. A multi-faceted study designed to evaluate relationships between different aseptic culture procedures and morphological origins of primary explants was carried out. Vegetative buds from various positions relative to the mother corm (defined by cardinal points on the compass) and explants from the male floral axis of ‘Maricongo’ (the ‘False-Horn’, or florally determinate type ‘progenitor’ of ‘Superplatano’), and ‘Superplatano’ (a ‘French-type’) were used as starting materials. Responses under field conditions were studied using a number of parameters including yield of commercially marketable fruits. We compared four populations of shoots, each of which derived from at least three different shoots from within one mat, shoots derived from vegetative and floral material from the same mat for both ‘Maricongo’ and ‘Superplatano’, and shoots derived from a number of floral buds of the same clone (‘Maricongo’) all of which were in culture for the same length of time. ‘Superplatano’ was stable whether from vegetative corm or floral bud apex. This shows conclusively that if the starting point in the micropropagation process is a stable Musa clone, our tissue culture procedure is reliable. Considerable variation in bunch phenotype was observed, however, in plants regenerated from ten of 12 shoot and floral meristems started from the ‘False-Horn’-type ‘Maricongo’. Change from ‘False-Horn’-type (determinate) to ‘French’-phenotype (indeterminate) was evident in each of the three locations. Frequency of bunch reversion varied from 0-4 to 100%, but was confined to individual originating stem tips rather than clones. The most dramatic bunch phenotypic change occurred in plants regenerated from clone 3. All plants regenerated from shoot 3-North bore 100% ‘French-type’ bunches. However, reversion in plants regenerated from shoot 3-West was only 0-8%, and no bunch phenotypic change was observed in plants from shoot 3-East. Plants regenerated from both shoot and male floral axis tips in ‘Maricongo’ clone 4 also bore ‘French-type’ bunches. Frequency of bunch reversion from shoot 4-East was 0-4% as compared to 2-6% from 4-fiorat. Bunch reversion occurred at the frequency of 2-0% when plants were regenerated from clone 6-floral. No bunch reversion was observed in plants regenerated from a single shoot tip in clones 1-West and 5-floral. No dwarfism was encountered in any of the tissue culture-derived plants. We conclude that tissue culture per se plays a very small role, if any, in the direct induction of off-types. Pre-existing characteristics of the primary explant determines whether products of a multiplication show fidelity or not. Our data suggests that ‘Maricongo’ is a chimera and that ‘Superplatano’ is a revertant off-type that results when breakdown of the chimera occurs. Large numbers of stable ‘Superplatano’ were produced from unstable ‘Maricongo’ and this affirms the value of micropropagation for generation of clones with desirable bunch phenotype. © 1993 Annals of Botany Company.
