Origins of the Australian Boab (Adansonia gregorii)


Claudia Vickers, Australian Institute of Biotechnology and Nanotechnology (AIBN),

Jack Pettigrew, Queensland Brain Institute (QBI),      The University of Queensland

DNA sequence from boabs (Adansonia gregorii) sampled across their distribution in the Kimberley could reveal both the site(s) of origin and the coalescence time. This information can be used to resolve the different scenarios for the arrival of this iconic tree in Australia. <> 


Baobab trees (Figure 1) attract attention because of their striking shape and the large number of different uses to which they can be put by human cultures (including food, water storage, medicine, raw materials for rope, cloth, twine, boats etc.). There are nine extant species of baobab (Adansonia spp.): six in Madagascar, two (A. digitata, A. kilima) on the African continent and one (A. gregorii) in the Kimberley region of north-western Australia. The Kimberley species is an obvious geographical outlier, and there are several different scenarios for the presence of the A. gregorii in Australia.


Malagasy  Baobabs.jpgOne theory proposes that Adansonia spp. share a common origin in Western Gondwana (Wickens 2008). However, recent molecular analysis has demonstrated that the Australian baobab, A. gregorii, is very closely related to the African species, A. digitata (Baum 1998). The genetic distance between the two species is far smaller than would be expected for a >100 Myr Gondwanan connection. This led to the suggestion that the baobab must have undergone transoceanic dispersal (Baum et al 1998, Pettigrew et al. 2014), presumably via floating seed pods. However, this theory is mechanistically constrained by several factors: (a) A. gregorii seed pod has the thinnest shell of all Adansonia spp., making it unlikely that seed pods would survive such a long oceanic journey (b) oceanic currents are unfavourable for the observed dispersal pattern and (c) A. gregorii is not present at other locations on the North-West Australian coast where it would readily grow and where oceanic dispersal would be expected to have delivered seeds.


We are investigating a third scenario based on the very close genetic relationship between A. gregorii and the African species: transoceanic dispersal mediated by human migrations out of Africa around 60-70,000 yrs ago. Interestingly, the geographical distribution of the Kimberley species overlaps almost perfectly with a particular type of ancient rock art known as Bradshaw paintings. The aetiology of these painting is under hot debate: some maintain that they are part of the extensive Aboriginal rock art found across Australia, and some maintain that these images were painted by a distinct culture which no longer survives in Australia. Bradshaw rock art is significantly different from other rock art in Australia in terms of style, materials used and the presnce of a biofilm of pigmented micro-organisms that define the contours and ensure an indefinite life. Scenes from daily life are strikingly well executed and fauna is very accurately depicted (Pettigrew et  al 2008). Moreover, there are many references in art that support a relationship between the artists, the baobab trees and intercontinental travel, for example: (1) the fruit and flowers of baobabs appear to be well represented in the images (2) large boats are featured in some paintings; these boats carry up to 30 passengers and have a very high prow indicating oceanic capability; they also have longitudinal striations like Thor Heyerdahl¹s ³Kon Tiki² and ³Ra² that could be boab fibres rather than reeds. One of the most striking things about these paintings is that they are remarkably reminiscent of modern African culture. It seems very likely that a stone-age African oceanic migrant might have chosen such a useful cargo as baobab, with its two dozen separate roles (Appendix), not to mention the durability of the nutritious fruit (the nutritious seeds surrounded by Vitamin C-rich pulp last for years).


If this species was brought to Australia by a culture from Africa, it should be possible to identify an ancestral population which should be found in a coastal location; if not, genetic diversity should be relatively similar across the Kimberly (Phase I). If we can identify an ancestral coastal population in Australia, we may then be able to identify the source of the ancestral baobab population by sampling from A. digitata populations in Africa (Phase II).



Using a 10 km grid, we have collected leaf samples from 250 A. gregorii trees across the Kimberley for analysis in Phase I of this project. Included in these samples are leaves from a number of island-located trees; these might enable time calibration of a haplotype network based on the 11 Kyr period since island-mainland divergence was driven by flooding at the end of the last glacial.

Haplotype Analysis: After DNA extraction using Qiagen Plant Kit, a number of different approaches will be used to explore intra-specific variation in A. gregorii:

    a. AFLP using multiple primers and fluorescent tags
  1. b. Sequencing of nITS gene.
  2. c. Sequencing of ³fast-moving² chloroplastic genes such matK
  3. d. Microsatellite DNA analysis    

These analyses will be performed on a sub-sample of 10 leaves to determine the most appropriate method for identifying intraspecific variation. Once the best technique has been identified, all samples will be analysed, any population characteristics will be identified and a haplotype network constructed and dated.


Progress Report:
    Although the leaf collection and gene sequencing of 220 boabs has been completed for years now, a final date for the boab's arrival has eluded us because of a dispute about mutation rate.

     The nITS sequences give us a non-controversial phylogenetic "tree", which has remarkable genetic uniformity, consistent with the recent arrival of of one, or a small number, of seeds. To use this phylogenetic tree to arrive at a date, one needs the mutation rate, which is not presently available for this species.
    One way around this mutation rate problem is to assume that the boab has the same mutation rate as its the sister African taxon with an accepted rate (A. digitata). Although this assumption of equality has been the basis of previous studies, it is risky. For example, plants in the same genus are known to have wildly different mutation rates for nITS. In addition, the boab lives in a completely different, extreme, niche compared to that of its African sisters, and has different pollination. When applied to the nITS data, this assumption gives a date of 2.4 million years, much too long ago to involve transport by humans.
    Alternatively, phylogenetic analysis of DNA sequence from nITS in all 9 baobab species shows that the boab is an outlier that has changed much more that the other 8 species over the same time; i.e. it has a much higher mutation rate, a conclusion quantitatively supported by an independent study of time dependence of mutation rate in this recent species (even 2.4 Ma is "recent" in terms of time dependence of mutation rate). This alternative estimate gives a greatly-inceased mutation rate that is high, but still within the 50X range of varation for the nITS gene. The increased rate gives an age for the boab of ~70 Ka. This is compatible with human transport, unlike the much longer estimate that results from the assumption of an African rate.
    Microsatellite DNA analysis also reveals "young genetics", with low variation also supporting a very recent arrival of A. gregorii. Despite the nITS substitutions' supporting a recent arrival, the assumption about equal Australian and African mutation rates tends to trump the data indicating a higher mutation rate in the boab. The discrepancy is handled by the invention of a "Pleistocene refugium", to explain the absence of expected boabs with 2.4 Ma-old DNA, a mystery wrapped in an enigma (Bell et al 2014).

Baum DA, Small RL,Wendel JF, 1998 Biogeography and Floral Evolution of Baobabs (Adansonia, Bombacaceae) as Inferred from Multiple Data Sets: Systematic Biology, 47: 181-207

Bell KL , Rangan,H, Fowler R, Kull CA, Pettigrew JD, Vickers CE , and Murphy DJ (2014) Genetic diversity and biogeography of the boab Adansonia gregorii (Malvaceae: Bombacoideae) Australian Journal of Botany 62(2) 164-174

Pettigrew JD, Nugent M,  McPhee A,Wallman J, (2008)  An Unexpected, Stripe-faced Flying Fox in Ice Age Rock Art of Australia¹s Kimberley. Antiquity 82: #318 December

Pettigrew JD. Bell, KL; Bhagwandin, A; Grinan, E; Jillani N; Meyer, J; Wabuyele, E; Vickers, CE. Morphology, ploidy and molecular phylogenetics reveal a new diploid species from Africa in the baobab genus Adansonia (Malvaceae: Bombacoideae) (2012) Taxon 61:1240-1250 (11)B

Wickens GE & Lowe P  2008 Baobabs - Pachycauls of Africa, Madagascar & Australia
Springer Verlag

 Baum DA, Small RL,Wendel JF, 1998 Biogeography and Floral Evolution of Baobabs (Adansonia, Bombacaceae) as Inferred from Multiple Data Sets: Systematic Biology, 47: 181-207

Pettigrew JD, Nugent M,  McPhee A,Wallman J, (2008)  An Unexpected, Stripe-faced Flying Fox in Ice Age Rock Art of Australia¹s Kimberley. Antiquity 82: #318 December

Wickens GE & Lowe P  2008 Baobabs - Pachycauls of Africa, Madagascar & Australia"
Springer Verlag B

Appendix: Utility of Baobabs


Water: Storage in hollowed tree; tap from trunk; masticate sub-cortical fibres.



. Leaves: salad when green; dried and ground addition to coucous; infusion.

. Tap root: sliced in salad.

. Fruit: Seeds give edible oil, ground meal; eaten whole if roasted.

            Pulp is highly nutritious and high in Vitamin C.

            (NB. fruit will keep for over a year in its convenient packaging, to suit a long voyage).

.Trunk and roots: Edible; can be completely consumed by a cow or elephant.



.Leaves: Infusion for cosmetic use, for gastrointestinal upsets, rickets etc





.Pollen: Flowers have abundant pollen that can be used to make glue.

.Fibres: Twine, nets, cloth, rope (very high quality with a performance comparable to nylon), boats (a la Ra of Thor Heyerdahl).

ŒPod Shell: Bowls, balers, merakas, breast covers.



            Hollow trees used as jails, garages, shops, bars, music rooms etc


Art and Religion:

            Roots: Red dye.

            Pod Shell: Ceremonial oval object; decorated.

            Fruit pulp: Organic acids (e.g. malic) used as mordants for inks

            Hollow tree: Entombment of deceased from selected professions.

            Base of tree: Meeting place for worship and discussion; campfires


Longevity and Resilience:

            Baobabs may have played a role in the survival of cultures that fostered them             because of the trees¹ ability to carry through a catastrophe (such as the Toba           event), when other sources of food might have disappeared.