SPORE MORPHOLOGY IN THE DICKSONIACEAE. I. THE GENERA CYSTODIUM,THYRSOPTERIS, AND CULCITA

The most mature spores available in herbarium specimens of the dicksoniaceous genera Cystodium, Thyrsopteris, and Culcita were studied by scanning electron microscopy and light microscopy, and representative specimens were analyzed to determine the number of spores produced per sporangium. Thyrsopteris and Culcita feature 64-spored sporangia, but Cystodium is consistently 32-spored. Spores were analyzed both in their native state as found on the specimens and in a perine-free state achieved by treatment with sodium hydroxide or acetolysis mixture. The sodium hydroxide assay demonstrated the presence of a perine in Cystodium, Thyrsopteris, and Culcita subgenus Culcita, but no evidence of a perine so defined was found in Culcita subgenus Calochlaena. Spores of Cystodium feature a nearly psilate exine overlain by a striate inner perine and a granular outer perine and are in several respects similar to those of Metaxya in the cyatheoid-dicksonioid complex and to those of Saccoloma in the dennstaedtioids. The most mature Thyrsopteris spores available may not have been fully mature. They feature a sparsely distributed, granular perinous layer over a microverrucate sculptine. The latter topography is taken as largely perinous since treatment with sodium hydroxide left a nearly psilate exine. The spore morphologies of Cystodium and Thyrsopteris reinforce the taxonomic distinctness of these monotypic genera indicated by their other characters. The two subgenera of Culcita are very dissimilar in their spore morphologies. The exine in subgenus Culcita ranges from psilate to slightly microverrucate proximally and distally, with varying margo development. Spores of subgenus Calochlaena are strongly differentiated from those of subgenus Culcita by their exine of broad spinules which vary in their degree of lateral fusion to each other and in the granular appearance of their distal surfaces. Spore morphology in Culcita strongly supports the argument of those who would raise its subgenera to generic rank.