Field of Science

Cunoniaceae and Friends

The Albany pitcher plant Cephalotus follicularis of south-western Australia. The streaked colours on the inside of the lid attract insects into the pitcher; the incurved teeth around the rim stop them from climbing back out. Photo by Holger Hennern.

The plant order Cunoniales was first established in 1926 to include plants with similar flowers to the Saxifragales but that were primarily woody rather than herbaceous (Dickison, 1975). Woodiness vs. herbaceousness is no longer considered that significant a feature in plant classification (sometimes you can find both in the same genus) and the content of the order has varied between classifications*. Today, the type family Cunoniaceae is included in the order Oxalidales and a taxon "Cunoniales" is no longer used as such. However, one of the two basal clades within the Oxalidales includes the Cunoniaceae and two ex-Cunoniales genera placed in their own families, Cephalotus and Brunellia, together with the family Elaeocarpaceae (previously in its own order) (Matthews & Endress, 2002). With the notable exception of Cephalotus, the members of this clade are mostly shrubs or trees. Economically, the clade is not overly significant: some species are used for wood; some have good reputations as honey sources for bees; a few produce edible fruits but do not appear to have been systematically cultivated for them. Members of all families bear their flowers clustered into (most often cymose) inflorescences; the size of individual flowers in the inflorescences varies between species. Brunellia and Cephalotus both produce flowers with thick sepals and no petals; in the other two families, petals may be present or absent (Matthews & Endress, 2002).

*As have most flowering plant "orders". There's a reason why order-level taxa don't get much day-to-day use among botanists compared to families.

Coachwood, Ceratopetalum apetalum, a member of the Cunoniaceae from eastern Australia. Photo by Melburnian.

The clade formed by these four families has a distinctly southern distribution in southern Africa, South and Central America, south-east Asia, Australia and New Zealand. Elaeocarpaceae are absent from continental Africa but are present in Madagascar. Also, while currently absent from India, they have been recorded from the fossil record there (Crayn et al., 2006). The Cunoniaceae include about 300 species, half in the genus Weinmannia, whereas the Elaeocarpaceae include about 600 species. Molecular studies have shown that an Australian radiation of dry-habitat shrubs previously regarded as the family Tremandraceae is in fact a subclade of the otherwise mostly rainforest-inhabiting Elaeocarpaceae (Crayn et al., 2006). Interestingly, the molecular dating study by Crayn et al. (2006) suggests that the 'Tremandraceae' developed scleromorphy (a suite of adaptations such as hardened leaves that are usually associated with arid habitats) some time before the Australian continent developed its current arid climate. While this may seem counter-intuitive, it is worth pointing out that the fossil record supports the same thing in the evolution of the genus Banksia (Mast & Givnish, 2002). It has been suggested that scleromorphy in these groups was therefore not originally an adaptation for arid living, but for growing in the poor soils of the Palaeogene Australian rainforest.

Prima donna, Elaeocarpus reticulatus. Fringed petals are characteristic of a number of species of Elaeocarpaceae. Elaeocarpaceae flowers are also adapted for buzz-pollination, where anthers do not release their pollen until vibrated by the wingbeats of their pollinating bees. Photo from Kate's Photo Diary.

The oddest member of this clade, however, has to be the Albany pitcher plant, Cephalotus follicularis. Restricted to the south-west corner of Western Australia, this plant bears a superficial resemblance to pitcher plants from other parts of the world, the mostly Indomalayan Nepenthaceae and the North American Sarraceniaceae. All three of these families have evolved the pitcher morphology independently, and can in fact be placed in separate subclasses: Cephalotus in the Rosidae, the Nepenthaceae in the Caryophyllidae and the Sarraceniaceae in the Asteridae. In the Nepenthaceae, the pitchers are developed from trendrils at the ends of the leaves; in the other two families, the entire leaves form the pitchers growing from a ground-level rhizome. In Cephalotus, only the outer leaves of an individual plant are developed into pitchers; the inner leaves remain flat and simple.


Crayn, D. M., M. Rossetto & D. J. Maynard. 2006. Molecular phylogeny and dating reveals an Oligo-Miocene radiation of dry-adapted shrubs (former Tremandraceae) from rainforest tree progenitors (Elaeocarpaceae) in Australia. American Journal of Botany 93 (9): 1328-1342.

Dickison, W. C. 1975. Studies on the floral anatomy of the Cunoniaceae. American Journal of Botany 62 (5): 433-447.

Mast, A. R., & T. J. Givnish. 2002. Historical biogeography and the origin of stomatal distributions in Banksia and Dryandra (Proteaceae) based on their cpDNA phylogeny. American Journal of Botany 89 (8): 1311-1323.

Matthews, M. L., & P. K. Endress. 2002. Comparative floral structure and systematics in Oxalidales (Oxalidaceae, Connaraceae, Brunelliaceae, Cephalotaceae, Cunoniaceae, Elaeocarpaceae, Tremandraceae). Botanical Journal of the Linnean Society 140 (4): 321-381.


  1. Wow, who woulda thunk that the coachwood and the albany pitcher plant would be close relatives? I find my intuitive sense of 'who is related to who' within metazoa to be pretty good most of the time, but with plants its hopeless. So what sort of ghost lineage is attached to Cephalotus? Is it deeply nested within this clade or a basal branch?

  2. I don't think there's a well-established phylogeny for relationships between the four families, but no-one has ever suggested that Cephalotus may be nested within Cunoniaceae or Elaeocarpaceae so either it (or possibly it + Brunellia) is quite likely to have a ghost lineage at least as old as their earliest records (some time in the Palaeogene).

  3. Hi Christopher.
    To me, as a casual plant observer, (well, casual observer of rainforest plants), the Cunoniaceae is a rag-bag, dumping basket of plants the Botanists could not come to terms with.
    It does not have the consistency of characteristics which one hopes for in a "good" botanical classification.
    I know the problem is in our perception of order in Nature, but somehow I think it shows we are "missing the point" somewhere along the line.
    If we aim to trace back relationships, in an evolutionary sense, the family relationships ought be more direct and simple. Instead of that, we end up with seemingly nonsensical "links".
    That simply makes me question the validity of the claims of family linkages. Time for a re-think I suspect.
    Denis Wilson

  4. On the contrary, the current Cunoniaceae is well supported by molecular data as long as the genera Davidsonia, Bauera and Eucryphia are included. Morphological support for the family is more difficult, though two synapomorphies proposed for the family by Bradford & Barnes (2001) interpetiolar stipules (subsequently become lateral in the tribe Geissoieae) and brachyparacytic stomata (subsequently altered in some Cunonieae). Brachyparacytic stomata are those where the stomata is bordered by two parallel subsidiary cells that do not reach fully to the ends of the stoma. Neither of these characters are unique to Cunoniaceae among flowering plants as a whole, but convergence is a major issue in plant evolution.

    Bradford, J. C., & R. W. Barnes. 2001. Phylogenetics and classification of Cunoniaceae (Oxalidales) using chloroplast DNA sequences and morphology. Systematic Botany 26 (2): 354-385.

  5. Beautiful pitcher plant! And strange! I am actually working on Sarracenia spp. in North Carolina right now. Characterizing the level of introgession within a hybrid zone between S. flava and S. purpurea. I must say getting DNA out of plants is pain in the ass. But its a cool system.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS