Composition | Position | Division
Malvaceae Info (Home)
The Fremontodendreae Pages
The tribe Fremontodendreae is composed of two genera. Chiranthodendron Sessé ex Larreat. is a monotypic genus from Guatemala and southern Mexico. Fremontodendron Coville has two or more species - there are several populations of uncertain rank - in northern Mexico, Arizona and California. The unity of the tribe is evidenced by shared derived characters, such as oppositifoliate flowers, the presence of only a single whorl of sterile floral parts, identified as the calyx (this apetalous trait is shared with Sterculieae, and with several other malvaceaous taxa including Cullenia, some species of Grewia, and some members of Lasiopetaleae), a quinquncial arrangement of the sepals, the arrangement of the stamens, and by the existence of inter-generic hybrids, and by DNA sequence evidence. The tribe also includes the intergeneric hybrid between these two genera, and the extinct genus Florissantia from the Palaeogene of western North America and the Russian Far East.
[Note: Professor Reveal has brought to my attention the facts that Chiranthodendreae Baill. (1872) had priority over Fremontodendreae Airy Shaw (1965), and that the rules of plant nomenclature do not at the present allow the conservation of tribal names. Therefore, under the correct application of the International Code of Botanical Nomenclature the tribe should be known as Chiranthodendreae. However the names Fremontieae and Fremontodendreae have long been used in the botanical literature instead of Chiranthodendreae. For the time being I am continuing to use Fremontodendreae.]
The tribe Fremontodendreae is relatively isolated morphologically. It has usually been placed in Sterculiaceae [5, 6], but Bentham [1, 2] placed it in Bombacaceae, and Asa Gray  removed it from Malvales altogether, erecting the family Cheiranthodendreae (Cheiranthodendraceae, Chiranthodendraceae). Bentham appears to have been correct, but perhaps not for the right reasons: he placed it within subtribe Matisiae of his tribe Bombaceae (he included Bombacaceae within Malvaceae), and molecular evidence places both Matisieae and Fremontodendreae within Malvatheca, and perhaps nearer to Malva than to Bombax. The correct division of Malvatheca is unclear - and it is possible that in the future it will be united into a single subfamily, or divided into three or four - but I place Fremontodendreae in sub-family Bombacoideae of Malvaceae sensu APG. (Kubitzki and Bayer  place both Matisieae and Fremontodendreae in sub-family Bombacoideae.)
Morphological evidence is unhelpful in positioning Fremontodendreae. The alternate, stipulate, petiolate foliage with palmate venation is a basal trait in Malvaceae. The loculicidal 5-valved capsules occur widely in Malvaceae, and may also be basal. The involucel of 3 free bracteoles is also widespread, and is probably basal to a clade including Malvatheca and other taxa. The stellate indumentum is also not useful, the indumentum being variable in Malvaceae, and not useful as a suprageneric character.
The quincuncial arrangement of the sepals is unique in Malvaceae, and therefore uninformative. (It was primarily this character which led Asa Gray to remove these genera from Malvaceae, placing it between Hypericales (now subsumed in Malpighiales) and Malvales; in other floral characters they do not agree with Hypericum.) The apetalous condition and oppositifoliate position of the flowers are both rare in Malvaceae, and none of their possessors is a candidate for the sister group of Fremontodendreae.
Fremontodendreae is atypical within Malvatheca in possessing dithecate anthers, and a staminal column divided about half-way into 5 stamens. Most taxa within Malvatheca have monothecate anthers and numerous stamens. Bentham interpreted the stamens as a fused pair of monothecate stamens, homologous to the clusters of monothecate stamens arising from each of the five lobes of the staminal column in Matisieae. In support of this interpretation he adduced the observed variation in the number (occasionally 3) of anther locules; however this is not compelling, as the number of locules of anthers is observed to vary, e.g. Megatritheca in Byttneroideae.
Fremontodendreae has pollen similar to that of some elements of Bombacoideae.
DNA sequence evidence is inconclusive. However Fremontodendreae lack a 2 residue insertion in the chloroplast matK gene product, which is present in the remainder of Malvatheca; this suggests that Fremontodendreae is sister to the remainder of Malvatheca . However sequence parsimony places Fremontodendreae equivocally closer to Malva than to Bombax.
Considering the absent matK insertion, the bombacoid pollen, its distribution, and the fossil record of 'bombacoids', Fremontodendreae can be hypothesised to be the sister group to the remainder of Malvatheca and the remnant of a Boreotropical 'bombacoid' radiation (crown group Malvatheca, excluding Fremontodendreae, is clearly a Neotropical clade). Alternatively it might represent a MesoAmerican dispersal within the Malvatheca crown group. In the absence of a clear solution I tentatively follow Kubitzki & Bayer is placing it within a possibly paraphyletic Bombacoideae.
As there are only two genera, and no doubts as to generic limits (except for proposals to unite the two genera in one) the division of the tribe is trivial.
Malvaceae Info (Home)
The Fremontodendreae Pages
© 2004, 2005, 2006, 2009 Stewart Robert Hinsley