Pollen ultrastructure of the Haemodoraceae and its taxonomic significance

The classification of the genera and tribes of the monocot family Haemodoraceae has been variable and uncertain. In order to provide additional taxonomic characters, SEhl and TEhl were used in describing the pollen ultrastructure of seventeen genera and nineteen species of the family. The Haemodoraceae as a whole, with the exception oflannria and Lophioka, is palynologically united by the absence of a typical tectate-columellate exine architecture. Within the family, all eight genera of the tribe Haemodoreae have boat-shaped, monosulcate, heteropdar pollen grains with a verrucate (foveolate in one genus) exine wall sculpturing. In contrast, all six genera of the tribe Conostylideae (as defined here) possess variously shaped 2-, 3-, or 7-8-porate, isopolar or apolar grains with a characteristic rugulate sculpturing. A distinctive 2-layered exine, lacking a footlayer, occurs in all members of the Conostylideae and in four genera of the Haemodoreae. The four other genera of the Haemodoreae are linked to the 2-layered members by a hypothesized gradation in wall architecture. Additional ultrastructural evidence supports the delimitation of the tribes Haemodoreae and Conostylideae, the placement of Phlebocarya within the tribe Conostylideae, and the recognition of subgroups of genera within each tribe. In contrast to the Haemodoraceae, members of the Tecophilaeaceae possess a tectate-columellate exine, arguing against the classification of that family as a third tribe (Conanthereae) of the Haemodoraceae. Lanaria and Lophiola are aberrant within the Haemodoraceae in having tectate-columellate exine mlls with monosulcate, exine-less apertures. The classification of Lanaria or Lophiola within the tribes Haemodoreae or Conostylideae or within the Tecophilaeaceae is refuted based on palynological evidence. The pollen of Pauridia, which also has a tectate-columellate architecture, is otherwise distinct from that of Lanaria and Lophiola in having disulcate apertures and in lacking an ektexinous foot-layer. The inclusion of Pauridia in the Haemodoraceae versus the Hypoxidaceae is unresolved based on the present study.