If the distribution of Malvaceae as a whole is considered, no clear patterns in distribution or habitat leap into view. The family is found in all continents, excepting Antartica, though predominantly in the warmer regions of the world. It is found in humid, semi-arid and arid habitats, the principal absence being from the polar, tundra and taiga zones. (Alpine species are confined to the Andes.) Species of Malvaceae range from ephemeral herbs to tall rain-forest trees.
Most clades and species are found in Mexico, tropical Central and South America, Africa, Madagascar, India, south east and east Asia, Papuasia and Australia. However the distribution reaches to the tips of South America (Cristaria in Tierra del Fuego), South Africa (Anisodontea and other taxa in the Cape Province), Australia (Asterotrichion in Tasmania) and New Zealand (Plagianthus on Stewart Island and the Chatham Islands). In the northern hemisphere the northerly limit is provided by Malva pusilla, which reaches 65°N in Europe. Several groups have species which reach relatively high latitudes, e.g. Malvastrum hispidum in the North American interior, Sidalcea hendersonii in British Columbia, Hibiscus syriacus in Korea and Hibiscus moscheutos in Ontario, Corchoropsis tomentosa in China and Korea and Grewia biloba in China. The lindens or limes, genus Tilia, are a predominantly cool temperate group, found as far north as Ontario, Finland and Manchuria,
There are a number of island endemic genera - Kokia and Hibiscadelphus in Hawaii, Acropogon and Maxwellia on New Caledonia, Pimia in Fiji, Hoheria, Plagianthus and Entelea in New Zealand, Lebronnecia in the Marquesas, Ungeria on Norfolk Island, Trochetiopsis on St. Helena, Astiria, Ruizia and Trochetia in the Mascarenes, and Tetralix and Neoregnellia on Cuba. Madagascar is host to a number of endemic genera, and is a centre of diversity for several others. Species of Malvaceae are also present on the Galapagos islands, in Macaronesia and the islands of the Mediterranean, the islands of the Caribbean, and are widespread in the islands of the central Pacific.
If one considers only the current distribution of Malvaceae it appears to be, with the exception of Tilioideae, to be by origin a southern hemisphere (Gondwanaland) group. However the current distribution of a taxon is a poor guide to its point of historical origin: one can't tell by distribution alone whether a disjunct distribution is caused by vicariance (the land moved), dispersal (the plants moved) or extinction (the plants used to exist in the gap). Further light can be shed by taking into account the fossil record (palaeobotany), the past distribution of land masses and climatic zones (palaeogeography) and the phylogeny of the plants (systematics). When this is done the southern hemisphere origin of Malvaceae is much less clear cut, and it looks as if the clade may not be so much southern hemisphere as tropical.
Be that as it may, the current distribution of the various elements of the family is as follows.
Byttneroideae is as a whole pantropical.
Within Byttneroideae Theobromeae is principally Neotropical, and tropical, with Guazama extending to the West Indies, and the single species of Glossostemon in arid regions of south west Asia. The New World taxa are forest trees.
Byttnerieae is pan-tropical, with all genera, except Byttneria itself, restricted to either the Old World or the New World. Byttneria is centred in the Americas, and the related Ayenia extends northwards into southern North America. Only a single species, Abroma augusta, is found in Australia, and this may be a relatively recent immigrant from Asia. Species in the tribe are shrubs, climbers and small trees of forest regions.
Lasiopetaleae is almost exclusively Australian (2 species of Keraudrenia are found in Madagascar, the aberrant Maxwellia lepidota is endemic to New Caledonia, Commersonia bartramii has a wide distribution from south east Asia through New Guinea and northern Australia to New Caledonia and Fiji, Commersonia novoguinensis is endemic to New Guinea, and Commersonia obliqua is endemic to Vanuatu. There are two clades; one is mostly Mediterranean, and especially diverse in south west Australia (but Maxwellia belongs to this clade), and the other tropical (but Seringia of south east Australia belongs to this clade). The majority of species of these clades occupy more or less xeric habitats.
In Hermannieae Hermannia is principally South African, but with 3 species in southern North America, 1 in Australia, 1 in Arabia and a few in Madagascar. Melochia is mostly American, but includes some palaeotropical species. Dicarpidium is Australian. Waltheria is centred in the Americas, especially Brasil and Mexico, but there are a few species endemic to Australia and Madagascar.
Grewioideae is also pantropical. Most genera are restricted to either the Old World or the New World (the exceptions are the pantropical Trichospermum, Corchorus and Triumfetta), and many of the Old World genera are restricted to either Africa or the Asia-Pacific region. Tetralix is endemic to Cuba, Pseudocorchorus to Madagascar, and Entelea to New Zealand. However there is no obvious pattern to the distribution above the level of genus, for example the nearest relative of the Afro-Madagascaran Sparrmannia is reckoned to be the New Zealand Entelea.
Brownlowiodieae is predominantly south east Asian (Sri Lanka and south east India, through south east Asia to the Pacific islands as far as Fiji, and north to Hainan, but Christiana has one species in Tahiti, two in South America, and a fourth ranging from Madagascar through tropical Africa to northern South America, and Carpodiptera is found in East Africa, the Comoros, Mexico, Trinidad and the Antilles.
Tilioideae are trees found in temperate decidouous forest in eastern North America (extending to south east Canada, north east Mexico and the highlands of southern Mexico), Europe and south west Asia, Siberia and east Asia, but absent from western North America and the temperate Himalayas. If Mortoniodendron is to be included here the range is extended to include the region between southern Mexico and northern Columbia.
Helicteroideae is pantropical.
Within Helicteroideae Durioneae is centred in Malesia, extending northwards to Burma, Thailand and the Philippines, and with Cullenia in Sri Lanka and southern India.
In Helictereae Reevesia is centred on continental south east and east Asia, but there are two species ("Veeresia") in Meso-America (a suspected occurrence in Bolivia turned out to be Pentaplaris (Malvoideae)).. Ungeria is an island endemic on Australia's Norfolk Island. Helicteres is found in the tropics of South America, Asia and Australia, but is absent from Africa; the related Neoregnellia is a Cuban endemic. Triplochiton is confined to Africa, and Mansonia is found in Africa and India.
Sterculioideae is centred in south east Asia, where the greatest generic diversity is to be found, but Cola and Octolobus are African, Acropogon New Caledonian, and Argyrodendron, Brachychiton and Franciscodendron Australasian. Sterculia, which contains over half the species in the subfamily is pantropical. Pterygota is also pantropical. Hildegardia has a discontinous pantropical distribution with species in Cuba, West and East Africa, Madagascar, India, Indonesia, the Philippines and Australia. Heretiera is predominantly south and south east Asian but extends to East Africa, Madagascar, Australia and the Pacific Islands. Firmiana and Scaphium extend from south east Asia to the Pacific Islands.
Dombeyoideae is palaeotropical, with centres of diversity in Madagascar and south east Asia. Melhania is the most widely distributed genus, being found in Africa, Madagascar, Asia and Australia. Trochetiopsis is endemic to St. Helena, Astiria, Ruizia and Trochetia to the Mascarenes, and Cheirolaena, Helmiopsis, Helmiopsiella and Paramelhania to Madagascar. Dombeya is centred in Madagascar, the Mascarenes and the Comoros, but extends to continental Africa and Arabia. Nesogordonia is Afro-Madagascarian. Harmsia is East African. Corchoropsis, Eriolaena and Paradombeya are centred in India and China, and Pterospermum, Schoutenia, Sicrea and Burretiodendron in south east Asia, whilst Pentapetes is widely distributed from India to the Philippines and northern Australia.
Bombacoideae and Malvoideae form a clade, Malvatheca, in which there are five lineages. Septotheca and Ochromeae are Neotropical. Fremontodendreae is found from Guatemala to California. Adansonieae is primarily Neotropical, but Adansonia is distributed in Madagascar, East Africa and north Australia, Bombax in tropical Africa, Asia and Australia, and Pachira is predominantly South American but extends to tropical Africa. Malvoideae is further divided into lineages. Matiseae, Pentaplaris and Uladendron are Neotropical. Radyera is found in Africa and Australia. The Lagunaria/Howittia/Camptostmenon group is found in eastern Australia, eastern Malesia and the Philippines. Alygoyne is Australian. Gossypieae and Hibisceae are pantropical. Malveae is centred in South America, but with lineages in the Old World, North America and Australia, and some widely distributed genera (Abutilon, Wissadula, Sida, Sidastrum).
The fossil record of Malvaceae extends back into the Cretaceous, but the group as a whole may be less than 100 million years old, perhaps considerably less, as the Rosid clade as a whole, by molecular clock estimates, is no older, and there are two substantial nodes (Eurosids II/Malvids and Malvales) between Rosids and Malvaceae. The crown group (the common ancestor of all living species and all its descendants) of Malvaceae probably originated in the Late Cretaceous, as apparently bombacoid and tilioid material has been found in deposits of that age, but it is not clear that the crown group of any of the subfamilies does so.
Of the 9 subfamilies of Malvaceae sensu lato Byttnerioideae and Grewioideae form a clade, which is the sister group to the remaining 7 subfamilies. The topology of the remaining families is not well resolved, but DNA sequence data favours Dombeyoideae as the sister group to the other 6.
The fossil record of Byttneroideae and Grewioideae is relatively sparse, and much of it not clearly placed within this clade. If the European leaf fossil Byttneriophyllum tiliaeofolium is correctly associated with the pollen fossil Intratriporopollenites instructus it is probably not byttneroid. The grewioid affinities of North American and European fossils of Grewia, Grewiopsis, Grewipollenites and Grewioxylon are also unclear. Consequently I discount them, and assess this clade as being of Gondwanan origin.
The distibution of Byttnerioideae can be interpreted as being due to a combination of vicariance (leading to tribes Lasiopetaleae in Australia, Hermannieae in south Africa and Theobromateae in South America) and subsequent dispersal events. Byttmerieae is pantropical, but its component groups might be the result of vicariance (Byttneria s.l. in South America, with later dispersal, Abroma in Malesia, and Scaphopetalum and Leptonychia in Africa. Glossostemon, which doesn't fit with the rest of Theobromateae in distribution or habitat, and Kleinhovia, which is morphologicallly anomalous in Byttnerioideae, may be particularly worthy of further study.
The fossil record of Grewioideae extends further north than the current range, Triumfetta being recorded from Yellowstone (Wyoming).
The crown group of Dombeyoideae seems to be of East Gondwanan origin. However dombeyoid wood is reported from the Eocene of North America (Chattawaya) and Europe (Dombeyoxylon), and fruits (Sphinxia) from Europe. Florissantia, from the Palaeogene of North America may also be dombeyoid, but I prefer to consider it as related to Fremontodendron and Chiranthodendron. Both the core of Dombeyoideae and Nesogordonia (which may be the sister group to the remainder of the subfamily) are centred on Madagascar. The presence of dombeyoids in south and south east Asia is plausibly the result of rafting on India, or on earlier blocks to cross the Tethys, or in the case of Melhania of later dispersal. As several genera (Schoutenia, Sicrea, Burretiodendron, Paradombeya) are specifically south east Asian transport on the Indo-China or Sibumasu blocks is a definite possibility. The presence of Corchoropsis in China, Korea and Japan may represent a case of dispersal.
Hibiscoxylon intertrappeum (possibly related to Pterospermum) is possible evidence of the early presence of dombeyoids in India.
The relationships between the remainuing subfamilies remains unclear, except that Bombacoideae and Malvoideae form a clade, and even after the transfer of the tribe Matisieae to Malvoidae the former may be paraphyletic with respect to the later. With the exception of Tilioideae the modern distribution of all these subfamilies suggests a southern origin. However, with the exception of Sterculioideae, the fossil record is not in accordance with this, and a common origin in the Boreo-Tropical flora, perhaps in North America, is a distinct possibility.
There are early records of Sterculioxylon in both Aftrica and Europe. (The fossil record of Colaxylon and Heritieroxylon accords with the modern distribution.) The wood of Sterculia is also recorded as a fossil from North America. However the current distribution suggests a southern origin. For example there is a distinct and appararently Australasian clade composed of Acropogon, Argyrodendron, Brachychiton and Franscicodendron.
The current centres of diversity of Tilioideae are in east Asia (Tilia and Craigia) and Central America (Mortoniodendron). However both Tilia and Craigia appear to have originated in North America, and have extensive fossil records throughout the northern hemisphere temperate zones. They may be viewed as a temperate offshoot of a Mortoniodendron-like member of the Boreo-Tropical flora of North America.
Cantitilia (Kentish Lime) is an early offshoot from the Eocene of England.
The fossil record of Helicteroideae is predominantly in the northern hemisphere. Fossil wood is known from North America (Triplochitioxylon) and east Asia (Reevesia and Wataria), The pollen genus Reevesiapollis has a long record in Europe, and is also found in North America, and Mansoniaepollis is an index fossil in the North Tethyan floral belt. I have found no reference to the fossil record of Helicteres. The current distribution of both Reevesia and Helicteres includes both the Oriental and Neotropical regions, but both are absent from Africa. Mansonia is found in Africa and India, and Triplochiton in Africa. These observations may be combined to infer an origin in the Boreo-Tropical flora, followed by subsequent expansion into the southern continents, and partial extinction in the north consequent on Neogene cooling. However, Ungeria, which on current distribution (endemic to Norfolk Island) might be assumed to represent a case of sweepstakes dispersal, has a significant pollen record (Reevesiapollis reticulatus) in Australia, and the earliest records of Durioneae (Lakiapollis) are from the Palaeocene of southern India
Brownlowioideae is currently concentrated in south east Asia. Although the pollen of tilioids, brownlowioids and some bombacoids is very similar, a pollen record apparently assignable to the Brownlowioids extends throughout the Tertiary of south east Asia. Otherwise fruit fossils are present in Europe (Christianocarpum) and pollen in the North Tethyan Belt (Tahitiaoipollis). Brownlowioideae might be unterpreted as a Boreo-Tropical element which early colonised south east Asia. However the distribution of Christiana (Tahiti, Africa, South America) and Carpodiptera (Africa, South America) questions this interpretation, especially as the fossils outside souith east Asia may well be closer to Christiana than to other genera.
The fossil record of Fremontodendreae is mostly restricted to western North America, where Florissantia is found as far north as British Columbia, but Florissantia is also recorded from the Russian Far East. The earliest records of other bombacoids are from North America (Bombacoxylon and Vulcanoxylon), but there is an extensive record of bombacoids from the Boreo-Tropical zone, with extensions as far as New Zealand.
The majority of fossil malvoids are found in South America, consistent with a South American origin for this clade. However there is an early record of Echinoperiporites from Pakistan.
© 2006, 2010 Stewart Robert Hinsley