Systematic anatomy of Oceanopapaver, a monotypic genus of the Gapparaceae from New Caledonia

Evidence from vegetative anatomy, reproductive morphology, and palynology does not support a relationship of Oceanopapaver with Cistaceae, Cruciferae, Flacourtiaceae, Papaveraceae, and Tiliaceae, but suggests placement of the genus in Capparaceae. The apparent occurrence of myrosin cells, among other features, effectively excludes all of the aforenoted families except Capparaceae and Cruciferae. However, multicellular non-glandular trichomes, bracteate inflorescences, sepals and petals each occasionally other than four per flower, presence of an androgynophore, numerous stamens, tricolporate and binucleate pollen, the unilocular mature ovary, the stipitate fruit, and the exotegmic seed in Oceanopapaver favour Capparaceae over Cruciferae. Floral histology and vasculature provide no clues about the relationships of Oceanopapaver. A few features are anomalous, most notably the presence of secretory canals and secretory cells in the genus versus their absence in Capparaceae and their rarity in Cruciferae, the trichomic floral nectary in the genus versus the massive, non-trichomic nectaries in these two families, and the straight embryo in the genus versus the more or less curved or folded embryo in the two families. The fleshy endosperm in Oceanopapaver has counterparts in a few Capparaceae, contrary to previous claims that endosperm is absent or scanty in this family. The report of stamen fascicle traces for Oceanopapaver is the first for Capparaceae, but these should be sought elsewhere in the family. Within Capparaceae the genus fits best in Capparoideae compared to Cleomoideae or the nine other very restrictive subfamilies variously proposed for Capparaceae. There is no justification for the monotypic segregate Oceanopapaveraceae. The phylogenetic and functional anatomy of vegetative and reproductive structures is discussed.