Since the time of Hallier (1905) and Bessey (1915), the origin of monocotyledons from a ranalean an cestor has been widely accepted (for a differing view see Burger 1981). On the basis of our increasing knowledge of the ranalean dicotyledons and their characters, Huber (1977) and others stressed the idea that the primary division in angiosperms is the one that separates the ranalean dicots plus the monocots from the remaining dicots ("eudicots"), a concept in part supported by molecular studies (e.g., Chase et al. 1993), although Ranunculales themselves are eudicots. The distinction between dicotyledons and monocotyledons is sometimes blurred, as Huber (1977) and Dahlgren et al. (1985) emphasized, by "dicot characters" appearing in the monocotyledons, and "monocot characters" occurring in dicotyledons. Dicot characters include an allegedly eustelic stem organization (Dioscoreaceae, Trichopodaceae), leaves with broad, net-veined blades (several groups, e.g., Dioscoreaceae and Convallariaceae), seed coats with a crystalliferous testa (Dioscoreaceae), embryos with terminal plumule (Dioscoreaceae, Taccaceae and others), seedlings with robust, long-lived and branched primary roots (Agave, Cordyline, palms, Strelitzia, Yucca; Tillich 1995), cellular endosperm formation (Acorus: Grayum 1987, 1991), and successive mi-crosporogenesis (some Asparagales, Arecaceae, and others: Rudall et al. 1997).
Character states that are common among monocotyledons but have a restricted representation in dicotyledons include trimerous flowers (e.g., Cabomba, Hedyosmum, Lactoris, and Saruma: Endress 1987, 1990, 1994.), adaxial prophylls (Aristolochiaceae, some Annonaceae, Lactorida-ceae), atactostelic stem organization (Nymphaea-ceae), Helobial endosperm formation (Cabomba), P2-subtype sieve-element plastids with cuneate protein crystalloids (Aristolochiaceae: Behnke 1976, 1981, 1995), and sulcate pollen. Among dicotyledons, these characters are largely restricted to magnolialean families, particularly Piperales, Nymphaeaceae, and Aristolochiaceae.
The scattered distribution of dicotlike characters in monocotyledons such as Dioscorea and Tacca led various authors, including Suessenguth (1921), to suggest a polyphyletic origin of monocotyledons from the magnoliids. However, analyses of both morphological and molecular data refute this theory. There remain some highly consistent synapomorphies for the monocotyledons, particularly P2-subtype sieve-element plastids (Behnke 1976, 1981, 1995) (Fig. 18), cotyledon number, and absence of vascular cambium.
The monophyly of the monocotyledons is well supported by several analyses, both morphological (Loconte and Stevenson 1991) and molecular, employing sequence data from rbcL (Chase et al. 1993, Qiu et al. 1993), rRNA (Hamby and Zimmer 1992, Soltis et al. 1997), and plastid DNA restriction sites (Davis 1995). These often indicate a group of primarily herbaceous taxa ("paleoherbs") including families such as Aristolochiaceae, Piperaceae, Chloranthaceae, and Nymphaeaceae as the closest relatives of the monocotyledons. However, all these groups are polymorphic with respect to plesiomorphic an-giosperm characters, and so far no single extant dicotyledonous family has been identified as the sister to the monocotyledons; indeed, their sister group may be a large clade containing many families, perhaps all magnoliids except for Nymphaeaceae and Winteraceae. Chase et al. (1995) and others have argued for inclusion of Magnoliales and Laurales along with paleoherbs for outgroup comparison. In many most parsimonious trees, either paleoherbs or Magnoliales and/or Laurales are the sister group to the monocotyledons. Since Magnoliales/Laurales do not include herbaceous groups, the question whether paleoherbs (including monocotyledons) were primarily herbaceous or diverged at an early stage from woody Magnoliales is not yet resolved, although there is little doubt that the immediate ancestor of the monocotyledons was herbaceous.
It is futile to search among extant dicotyledons for a monocot ancestor, because monocotyledons evolved from early angiosperms. Attempts to estimate the time of divergence by applying a molecular clock have yielded widely divergent results between a calculated age of 319 + 33 million years ago (Ma) (Martin et al. 1989) and 200Ma (Wolfe et al. 1989). Confidence in these estimates may be lessened, however, by the considerable differences in substitution rate between various plant groups (although most studies factor out this effect by performing relative rate tests and eliminating those groups that appear to be significantly faster or slower than average). Between grasses and palms, for instance, this amounts to a fivefold difference, a value consistent with the differences in generation time (Gaut et al. 1992). Palynological evidence from the lower Cretaceous of 140 Ma (Hauterivian of Israel, Brenner 1996) indicate the latest date for the di-cot/monocot split, which, of course, may extend back further.
l'ig. 18A-F. Sieve-element plastids of monocotyledons invariably are characterized by cuneate protein crystals (P2c type, llehnke 1981). This type is restricted to monocots (exception: Asarum, Aristolochiaceae). Although the pure P2c type is the most frequent form, additionally, starch grains (P2cs, in Dioscoreaceae, Araceae, Zingiberales, and palms) and/or peripheral protein filaments (P2cfs, in Musaceae and Monstera, l'2cf in Lomandraceae, Asphodelaceae) may be present A Af.orus calamus, P2c. B Gymnostachys anceps, P2cs. C Cala-ilium bicolor, P2cs. D Freycinetia cumingiana, P2cf. E Musa Sllmatrana, P2cfs. F Monstera deliciosa, P2cfs. TEM X2000. (Original H.-D. Behnke)
Was this article helpful?