Flower structure and development in Tupidanthus calyptratus (Araliaceae): an extreme case of polymery among asterids D. D. Sokoloff 1 , A. A. Oskolski 2 , M. V. Remizowa 1 , and M. S. Nuraliev 1 1 Department of Higher Plants, Biological Faculty, Moscow State University, Moscow, Russia 2 Botanical Museum, V.L. Komarov Botanical Institute of the Russian Academy of Sciences, St. Petersburg, Russia Received March 16, 2007; accepted April 16, 2007 Published online: July 2, 2007 Ó Springer-Verlag 2007 Abstract. Flowers of Tupidanthus show an extreme case of floral polymery among asterids. Floral development and gynoecium structure have been examined. The floral meristem has a complex folded shape. The tiny calyx is initiated as a continuous ring primordium. The corolla is initi- ated as a lobed ring and develops into a calyptra. All stamen primordia appear simultaneously as a single whorl. The carpels, also in a single whorl, tend to alternate with the stamens. Some Schefflera species related to Tupidanthus are also studied. The flower of Tupidanthus is interpreted as a result of fasciation. Further investigation should determine whether mutation(s) in gene(s) of the CLAVATA family are responsible for the fasciation here. The significance of Tupidanthus for understanding spa- tial pattern formation in flowers of Araliaceae, and both functional and developmental constraints in angiosperm flowers with a single polymerous carpel whorl are discussed. Key words: Araliaceae, development, evolution, fasciation, flower, polymery, Schefflera, Tupidanthus. Compared with other eudicots, asterids are generally characterized by relatively stable and synorganized flower construction (Endress 1990). Asterid flowers are almost exclusively whorled and typically pentamerous (except the gynoecium, which is often dimerous). How- ever, some asterid clades exhibit considerable meristic variation and increased number of floral parts, prompting much discussion regarding the possible primitive or derived nature of such polymerous flowers (or flowers with polymerous androecia and/or gynoecia), especially within Araliaceae (Apiales). Most recent authors support a secondary increase of flower merism as the most likely explanation of polymery in these anomalous asterids (reviewed by Ronse De Craene and Smets 1998 and Endress 2002). At least in the order Apiales, parsimonious mapping of flower mer- ism onto existing molecular phylogenetic trees strongly supports the derived nature of polymery (see below). Nevertheless, the likely derived nature of polymery in the Asterids does not reduce its impact on broad-scale discussions of flower evolution in angiosperms. Detailed investiga- tion of these anomalous species will help to elucidate the developmental and functional constraints that exist in asterid flowers in general. Studies of extreme forms help to Pl. Syst. Evol. 268: 209–234 (2007) DOI 10.1007/s00606-007-0559-5 Printed in The Netherlands Plant Systematics and Evolution