The fossil record of the Malvaceae sensu APG extends back as least as far as the Upper Cretaceous (Campanian and Maastrichtian), but most subfamilies are first recorded from the Palaeocene or Eocene. As the Upper Cretaceous fossils belong to the Bombacoideae, which is one of the younger groups within the family, it can be predicted that a more extensive Cretaceous fossil record is yet to be found or identified.
The richest fossil records are of bombacoid pollen, which is found in North America in the Cretaceous, is found throughout tropical regions during the Palaeogene, and is mostly confined to South America during the Neogene, and of Tilia for which a pollen and macrofossil record spans the Eocene to Pleistocene of the northern temperate regions.
Fossil material of plants can be difficult to identify as to species, genus, or larger taxonomic units, as usually what is found is individual parts of plants, such as wood, leaves, flowers, fruits or pollen, and these are often insufficient for identification, particularly for older material which is less closely related to modern material, and may be less well preserved. Consequently, and as fossils of one plant part often cannot be unambigiously associated with those of another plant part, palaeobotanists use form genera to classify parts of plants of uncertain taxonomic position.
These pages (will) list and give outline information on those fossil plants which are believed to be assignable to Malvaceae, and also, to avoid ambiguity, a number of plants which were previously thought to be malvaceous, but for which the current opinion differs.
Fossils can be found under the names of modern genera (including Abutilon, Bombax, Brachychiton, Burretiodendron, Catostemma, Craigia, Dombeya, Eriotheca, Firmiana, Gossypium, Grewia, Guazuma, Hampea, Helicteres, Hibiscus, Hoheria, Kleinhovia, Kydia, Luehea, Pterospermum, Reevesia, Sterculia, Tilia and Triumfetta).
The assignment of fossil plants to taxa has been an error-prone process; for example Nordenskioldia borealis belongs to Trochodendraceae, not Malvaceae, as once thought, Sterculiocarpus coloradensis to Papaveraceae, and many fossils originally placed in "Grewia" are really hamamelidaceous. In the reverse direction Pteleaecarpum and Florissantia were not originally recognised as malvaceous. I cannot be confident that several of the plants described here are indeed malvaceous
There are also possible points of confusion related to homonyms. At least 3 generic names of fossil plants are later homonyms of malvaceous genera, one being a fern, a second possible a fern, and the third probably a seed fern.
© 2005, 2006, 2008, 2009, 2010, 2012, 2020 Stewart R. Hinsley