The genera Mitchella and Damnacanthus. Evidence for their close alliance; comments on the campylotropy in the Rubiaceae and the circumscription of the Morindeae

The two species of Mitchella (Southeast Asian and North American) and several species of the Southeast Asian genus Damnacanthus are investigated. Vegetative character states (growth form, branching pattern, leaves) of the two genera are described and compared. Damnacanthus always exhibits heterophylly and some species have paired thorns. The latter are interpreted as paired lateral shoots in the proximal part of a sympodial branch unit, and it is speculated that the paired thorns may be modified inflorescence shoots. Detailed information is also given on inflorescence structure and floral morphological and anatomical details of Damnacanthus and Mitchella. Particular emphasis is placed on the ovaries, the structure of which is remarkably similar in the two genera. The uniovulate locules are characterized by having campylotropous ovules which are inserted near the top of the septum; extensive obturator tissue covers part of the horizontally arranged curved ovule in cap-like manner. The micropyle of the ovule, obscured by the obturator, points upwards and to some degree also inwards, while the embryo sac is found in a ± horizontal position. The seeds, however, contain minute embryos, of which the radicles are pointing ± downwards. This apparently contradicting micropyle and radicle position finds its explanation in the unusual ovule structure and orientation and in the subsequent strong growth of the endosperm, through which the embryo is pushed to the position where it is found in mature seeds. The detailed structure of the drupaceous fruits and of the seeds of both genera are compared. Chromosome numbers (x = n = 11, 2n = 22) are presented for Mitchella and Damnacanthus, certain palynological information is added, and literature on reproductive biological aspects is reviewed and supplemented by original observations. While Mitchella had been associated with various different tribes, Damnacanthus was in the past nearly always placed in the Morindeae. The investigated characters overwhelmingly support the close alliance between Mitchella and Damnacanthus. However, according to our present state of knowledge, the Morindeae appear to be a heterogeneous tribe, and at least a group of genera allied to Prismatomeris is probably to be removed. Damnacanthus and Mitchella do show certain agreements with the ‘core’ of the Morindeae, but their definite tribal placement should be withheld until a recircumscription of the Morindeae becomes available.