A PALYNOLOGICAL STUDYOF CROTONOIDEAE(EUPHORBIACEAE)1Joan W. Nowicke2ABSTRACT Pollen of 69 species representing 34 genera from 12 of the 13 tribes of Crotonoideae sensu Webster was examined in light microscopy (LM), scanning electron microscopy (SEM), and transmission electron microscopy (TEM). The pollen of the vast majority of the subfamily is inaperturate, and ail Crotonoideae exines have a similar architecture: triangular supratectal elements attached to a network of muri having short or irregular columellae. These three components, triangular supratectal elements, muri, and modified columellae constitute the Croton structure. The trangular elements, designated subunits, can be psilate, striate, furrowed and ridged, or pitted, and with echinate, ronwded, or long-attenuate apices; the subunits can be closely spaced or arrayed in more open configurations. A Croton sculpture refers to the presence of triangular sculpture elements, which may or may not have muri and modified columellae. Inaperturate grains have thin footlayers and thin endexines, whereas colpate pollen has both footlayer and endexine well developed; in porate grains, the footlayer and endexine appear intermediate between iaperturate and colpate. Tribes Micrandreae (Hevea, Micrandra, Micrandropsis examined) and Adenoclineae (Adenocline, Ditta, Glycydendron, Endospermum, Klaineanthus, Tetrorchidium) are 3-colpate and probably prim-aiie, having retained not only apertures but (prominent) footlayers and endexines as well; Manihoteae (Manihot, Cnidoscolus examined) and Gelonieae (Suregada) are porate. The thicker aperturate exines, especially in the colpate grains, may be linked with retention of apertures. The remaining 22 genera examined, including the two largest, roton and Jatropha, have inaperturate pollen. A thin footlayer/endexine also occurs in Oldfieldioideae, and a tectum with triangular elements is found in some Acalyphoideae. In both morphology and exine structure, the pollen ofCrtonoideae closely resembles the (porate) pollen of Thymelaeaceae. In a family as large and diverse as Euphorbi-aceae (300 genera, 7500 species), variation in ollen morphology/structure would be expected. ln fact, there is a striking array of pollen diversity as deonstrated by Punt's (1962) classic study of uphorbiaceae in which numerous pollen types are lustrated with line drawings. Even genera, e.g., hyllanthus (Punt, 1987) and Zimmermannia (Poole, 1981), can be highly variable in pollen morphology, thus casting some doubt on the tax-onomic value of pollen data in this family. In his 1962 study, Punt followed the treatmentf Euphorbiaceae in Pax & Hoffmann (1931), rec-ognzg only two subfamilies, Crotonoideae andhyllanthoideae Later, he (1987) consideredWebster's (1975) system that elevated three Cro-tonoideae tribes to subfamily status, Oldfieldioi-deae, Acalyphoideae, and Euphorbioideae, in ad-tion to Phyllanthoideae and the remainingrottoideae, to represent the best concept of thefamily to date. In part, pollen data support Web-ster's classification (see Discussion). By segregating the above three tribes from Cro-tonoideae, the remaining members of this subfamily form a much more natural alliance that may or may not be monophyletic. The Crotonoideae, sensu Webster, comprise 13 tribes and approximately 73 genera, of which three, Croton, Jatropha, and Manihot, account for half of the 2000 species. Crotonoideae now appear to be the exception to pollen diversity and have sur-prisingly uniform pollen morphology, united by the croton-pattern. This tern apparently originated with Erdtman (1952, 1966: 173) for pollen with a sexine that "consists of well defined, regularly ar-ranged excrescences (triangular or more or less circular in cross section), supported (always?) by a baculate, baculoidate, or spongy layer." He noted that most of the croton-pattern pollen was nonaper-turate and spheroidal. Punt (1962) examined at Sthank Carl Annable, Christine Begle, and Lura Wiliamson for technical assistance, Susann Braden and WalterS fr h with scanring electron microscopy, and Muriel Poston for helpful conments on the manuscript. Gradyr graciusly shared his scanning electron micrographs of Crotonoideae polen, and white only four cpees, A . n uborum, Deutzianthus tonkinensis, Pantadenia adenanthera, and Ricinocarpos stylosur appearath~je, the remainder provided background information. John Skvarla lent transmission electron micrographs of Stia lschaultii, Croton draco, C. texensis, and Joannesia princeps. yUS.PA.ent of Botany, National Museum of Natural History, Smrnthsonian Institution, Washington, D.C. 20560, ANN. MIsSOURI BOT. GARD. 81: 245-269. 1994.
BioStor
