Embryology of Japonolirion (Petrosaviaceae, Petrosaviales): a comparison with other monocots

Abstract

Japonolirion osense, the sole species of the genus, endemic to Japan, which is placed together with Petrosavia in the Petrosaviaceae and the order Petrosaviales, is still poorly known with respect to systematic characters. Here I present an embryological study of the anther, ovule, and seed of J. osense. Japonolirion is characterized by a glandular anther tapetum, simultaneous cytokinesis in the microspore mother cell, two-celled mature pollen grains, anatropous and crassinucellate ovules, a two-cell-layered nucellar cap formed early in ovule development, antipodal cells hypertrophied in post-fertilization stages, the ab initio cellular mode of endosperm formation, and exotegmic seeds. Comparisons with the basal monocots Acorus (Acorales) and Araceae (Alismatales), and with the more derived monocots Nartheciaceae (Dioscoreales) and Velloziaceae/Triuridaceae (Pandanales), showed that Japonolirion is clearly distinct from those basal and more derived monocots, supporting a distinct position for Petrosaviaceae or Petrosaviales within the monocots. Extensive comparisons further suggest that the two-cell-layered nucellar cap, whose cells are rich in cytoplasm at the time of fertilization in Japonolirion and thus obviously function as the obturator, is likely to be a common characteristic of the basal monocots and may even be a link with the magnoliids.

Embryology of Petrosavia (Petrosaviaceae, Petrosaviales): evidence for the distinctness of the family from other monocots

“…This study Tobe (2008) Buell (1935, 1938; Duvall (2001); Floyd and Friedman (2000); Jüssen (1929); Rudall and Furness (1997); Seubert (1993) Tobe and Kadokawa (2008, and references therein) Browne (1961); Sato and Kirito (1988); Stenar (1931); Varitchak (1940), Johri et al (1992 and unrelated monocot families, and even many dicot families. Thus, they do not indicate mutual associations among plants sharing the characters.…”

“…Although they possess many embryological features in common, Petrosavia and Japonolirion clearly differ in a few characters. For example, antipodal cells are ephemeral in Petrosavia but become hypertrophied after fertilization in Japonolirion; a two-cell-layered nucellar cap, with cells that become enriched in the cytoplasm just at the time of fertilization, forms in Japonolirion, but not in Petrosavia; the division of the primary endosperm nucleus results in one large micropylar and one small chalazal endosperm cell in Petrosavia but two equal-sized endosperm cells in Japonolirion (for Japonolirion embryological data, see Tobe 2008). Besides these embryological differences, the differences in habit (mycoheterotrophic in Petrosavia; autotrophic in Japonolirion), inflorescence (coryms in Petrosavia; racemes in Japonolirion), ovary position (semiinferior in Petrosavia; superior in Japonolirion), and chromosome number (x = 15 in Petrosavia; x = 12 or 13 in Japonolirion) also clearly distinguish the two genera from each other (for a summary of similarities and dissimilarities, see Cameron et al 2003).…”

“…Thus, comparisons must encompass a broad framework of monocot phylogeny. In a previous paper, Tobe (2008) compared Japonolirion with the basal monocots Acorales and Alismatales and the derived monocots Dioscoreales and Pandanales. Here, we compare the Petrosaviaceae with these groups.…”

“…Within the Araceae, we chose Gymnostachys, Symplocarpus, Lysichiton, and Orontium, which constitute a basally diverging clade in Araceae phylogeny Tam et al 2004). Tobe (2008) adopted this as the best method for comparing Japonolirion with the Alismatales. As is the rest of the family, the basal Araceae (Gymnostachys, Symplocarpus, Lysichiton, and Orontium) are poorly known embryologically, but fragmentary data are available (for basal Araceae embryology, see the review by Tobe and Kadokawa 2008 and references therein; see also Table 1).…”

“…Petrosaviaceae hold important systematic and evolutionary position, and in view of this, Tobe (2008) initiated the first thorough study of the family with an investigation of J. osense embryology. He demonstrated that Japonolirion is clearly distinct from both basal (Acorales and Alismatales) and more derived monocots (Dioscoreales and Pandanales).…”