Studentthree

ASTERIDS - Sympetalae redux?: (iridoids +); C sympetalous, if only early in development, petals largely free, (nectary gynoecial), ovules unitegmic, integument thick, endothelium +, nucellar epidermis does not persist, style single, long; seed exotestal; endosperm cellular, embryo long. [Back to Index]

The age of the stem group asterids may be ca 128 million years before present, mid Early Cretaceous, the Cornales and Ericales diverging soon afterwards, and the other asterid orders all diverging over 100 million years before present (K. Bremer et al. 2004); Anderson et al. (2005: asterids other than Cornales and Ericales not sampled) suggest figures of ca 112 million years before present for the stem group, ca 109 million years before present for the crown group. Iridoids, common in asterids, have been implicated in herbivore preferences, detering some and attracting others (e.g. see discussion under Plantaginaceae, Scrophulariaceae, etc.: Bowers 1980, 1988). Iridoids have a bitter taste and are emetics for vertebrates, at least, yet iridoids are sometimes sequestered by the insect and used in its defence against predators (Nishida 2002 for a summary). Thus Uraniidae (moths) are found on Dipsacales, Lamiales, Gentianales - and also Daphniphyllaceae, one of the few iridoid-containing members of Saxifragales (see Lees & Smith 1991). Larvae of Nymphalidae-Melitaeini butterflies are also almost restricted to asterids. However, in this case they are also quite common on Asteraceae and Acanthaceae, which, although asterids, lack iridoids, furthermore, Melitaeini distinguish between plants with route I secoiridoids, which they commonly eat, and those with route II decarboxylated iridoids (iridoid glycosides), which they rarely eat (Wahlberg 2001).

In the asterids, compound leaves are relatively uncommon, and the leaflets are often not articulated and/or distinct. Taxa with stipules are also fairly uncommon, as are those with apetalous flowers, complex seed coats, and arillate seeds. Within asterids, Ericales and Cornales in particular show much variation in the degree of sympetaly, stamen number, adnation of stamens to corolla, ovule morphology and anatomy, and seed coat anatomy, etc. Some of this variation is like that found in rosids, Dilleniales, etc., and unlike that in the asterid I + II group, which see for more details. They may also have ellagic acid, which has a rather similar distribution.

Some families included in the asterids seem to be polypetalous, but developmental studies like those of Erbar (1991) may show they have a ring primordium very early on (see also, for example, Reidt & Leins 1994). However, the position of "type" of corolla initiation on the tree is quite uncertain. Apiales + Asterales + Dipsacales of the asterid II group have many members with early tube initiation, as do both Oleaceae and Rubiaceae, "basal" or almost so in their orders in asterid I group, and so do some Cornales. Although sampling leaves a great deal to be desired, early initiation could conceivably be a synapomorphy for the asterids as a whole (see Leins & Erbar 2003b for details). Whatever the method of tube initiation, it is only in Ericales and other asterids that the mature flower may have an obvious corolla tube, hence the apomorphy of corolla tube uniting [Ericales + other asterids]. Note that a tube-forming hypanthium is rare in asterids, although in the largely polypetalous rosids a hypanthium is quite common. In general, where many characters are to be placed on the tree depends on resolution of relationships within Ericales and Cornales, and the pattern of gain-loss of some of these features is liable to be rather complex.

Some characters common in asterids may have functional and logical linkages that must be taken into account. Thus the presence of a tenuinucellate nucellus is linked with that of unitegmic ovules, the development of an endothelium (Kapil & Tiwari 1978), and a simple exotestal seed type (Netolitzky 1926); that of sympetalous monosymmetric flowers with epipetalous stamens, etc.

The monophyly of the asterids is well established (e.g. Olmstead et al. 1992, 2000; P. Soltis 1999); Albach et al. (1998) suggest the four main groupings recognised here. A phylogeny proposed by K. Bremer et al. (2001) is based on an analysis of 2 genes + morphology, one by Albach et al. (2001b) on four genes; relationships suggested by these studies are largely congruent, differences being almost entirely in taxa not assigned to orders by A.P.G. (1998); see also the Bayesian analyses of Lundberg (2001b, d). B. Bremer et al. (2003) provide a recent comprehensive phylogeny of the clade, although with minimal sampling within families, using three coding and three non-coding chloroplast markers. Both B. Bremer et al. (2003) and Olmstead (2000) found strong support for Cornales being sister to all other asterids (see also Albach et al. 2001; Soltis et al. (2003). However, in Hilu et al. (2003: matK gene alone) the positions of Cornales and Ericales were reversed. Note that in their study, Caiophora (Loasaceae) appears in Asterales, far separate from the other members of the family in the analysis - perhaps mistaken identity?

Sympetalae of older studies were defined largely by their sympetalous corolla, and so families like Pittosporaceae,