Origin and evolutionary dynamics of Australian Elaeocarpaceae

A project undertaken at the Royal Botanic Gardens and Domain Trust, Sydney, and supervised by M. Rossetto and D. Crayn.


How and why did some representatives of the original Gondwanan flora survive within the small remaining pockets of Australian rainforest, while others disappeared or adapted to arid conditions and radiated more widely? This project aims to investigate the phylogeny, biogeography and within-species diversity of the plant family Elaeocarpaceae in order to understand some of the evolutionary mechanisms that have influenced speciation and distribution patterns within the Australian flora.
Australian rainforests provide a unique model for the investigation of plant evolutionary patterns. Despite representing less than 0.5 % of the total vegetation cover, it has been estimated that Australian rainforests hold up to 50 % of our terrestrial biodiversity. No other rainforest worldwide contains such a high number of rare ancient taxa, a consequence of the highly distinct evolutionary history of Australian rainforest. After 30 million years of isolation and stability, drastic climatic changes beginning approximately 15 million years ago introduced arid conditions in what was a predominantly rainforest dominated continent. These changes were further intensified during the glacial events of the Quaternary contributing to the uniqueness of our native flora.
The Elaeocarpaceae are a worldwide family of over 500 species across nine genera. Members of the family are trees or shrubs, usually found in rainforest environments. The major centres of diversity are in the Indo-Malayan and Australia-Pacific regions. Seven genera are only found in Australia if the endemic family Tremandraceae is included as suggested by recent studies.  


This project aims to investigate two broad themes on the biogeography and evolution of the Elaeocarpaceae. The first theme explored phylogenetic, biogeographic and evolutionary relationships among the Australian representatives of the family based on between-species DNA differentiation. Based on this phylogenetic framework, a second set of questions investigated in more detail the factors influencing the distribution of rainforest trees by measuring within-species differentiation in a representative genus, Elaeocarpus. Answers to these questions provided insights on how ecological traits, evolutionary potential, biogeographic origin and post-glacial survival have shaped current floristic distributions in Australian rainforests. Understanding the evolutionary basis that enabled the perseverance of taxa despite extreme environmental pressures will help us develop better conservation and management strategies for the future.

Main findings

Phylogenetics and evolution of Elaeocarpaceae – Divergence time analysis using molecular phylogenies and fossil-calibrated evolutionary rates can distinguish between competing historical biogeographic scenarios (e.g. dispersal and vicariance). In a detailed study of Elaeocarpaceae sensu lato (including Tremandraceae) we used plastid trnL-trnF and nuclear ITS sequence data and parsimony and Bayesian methods to demonstrate the monophyly of all recognised genera, infer their relationships, and estimate the minimum divergence dates (Crayn et al. 2006). The Tremandraceae clade, which consists of three genera and about 49 species of shrubby, dry-adapted Australian plants was nested within the widespread predominantly rainforest tree family Elaeocarpaceae (nine genera, over 500 species) as sister to a clade comprising Aceratium, Elaeocarpus and Sericolea. These two lineages diverged during the Palaeocene, after which the evolutionary rate accelerated markedly in former Tremandraceae. Extant members of former Tremandraceae trace their origins to the Oligocene but the major diversification occurred during the late Miocene, when environments in Australia underwent progressive aridification accompanied by rapid diversification in several sclerophyllous groups. The importance of dispersal in explaining the current geographical distribution of this family was illustrated by Aristotelia. The distribution of this genus, previously thought to be consistent with Gondwanan vicariance, was shown to be a result of dispersal, the ancestor of the two New Zealand species arriving from Australia at least 6–7 Ma. 

Why is a once-widespread lineage now confined to a narrow distribution?  –To assess the status of a putative new species Elaeocarpus L. (Elaeocarpaceae) from northeastern New South Wales (NSW) with respect to the morphologically similar E. blepharoceras Schltr. from New Guinea, we undertook morphometric analysis of 11 vegetative attributes measured on 11 specimens of the putative new species and nine of E. blepharoceras (Maynard et al. 2009). Cluster analysis (Flexible UPGMA) and ordination (PCC) separated highland specimens of E. blepharoceras from the NSW material plus lowland E. blepharoceras. Furthermore, the ordination showed some separation of the NSW material into Koonyum Range and Nightcap Range groups. Although it is not clearly differentiated from lowland E. blepharoceras on morphometric analysis, description of the NSW material as E. sedentarius D.J.Maynard & Crayn was justified by a number of morphological and distributional factors. Nevertheless the question remained as to why E. sedentarius has such a narrow current distribution. 
To clarify the biogeographic and evolutionary history of this lineage, we needed to understand if this highly endemic taxa is at the early stages of expansion; nearing the end of a period of decline; or persisting at low numbers over the long term? We combined molecular, environmental and ecological data to identify the factors responsible for the narrow distribution of a paleo-endemic rainforest tree: Elaeocarpus sedentarius (Rossetto et al. 2008). Between-population and between-generation comparisons of genetic diversity across all known populations of E. sedentarius showed evidence of mutation-drift equilibrium rather than evidence of a recent bottleneck. Similarly, floristic and environmental data negated the hypothesis of rarity as a consequence of highly specialized habitat requirements. Instead, genetic structure and the available ecological data supported the hypothesis of dispersal limitation as the main cause of endemism, and that the species may have attained genetic equilibrium without realizing its full niche potential. We suggest that these factors are likely to explain narrow endemism in a broader range of rainforest trees as well as other narrow endemic across the continent.

The use of multi-species approaches to discover the current and historical factors shaping the distribution of rainforest trees – Rainforest contractions caused by the aridification of the continent and the recent glacial cycles have left discrete genetic signatures on modern-day populations of rainforest trees. The nature of between-population differentiation is likely to have been influenced by a range of ecological and environmental factors. We used microsatellites to examine range-wide population genetic structure in two congeneric rainforest trees, Elaeocarpus angustifolius and E. largiflorens (Rossetto et al. 2007), with similar habitat preference and dispersal potential. The aim was to investigate the relationships between genetic structure, geographic disjunction and morphological differentiation and attempt to clarify the likely evolutionary processes responsible for the observed patterns. We found substantial differences in the amount and type of genetic differentiation within the two co-distributed species. E. largiflorens revealed an abrupt genetic-disjunction front between two subspecies separated by a biogeographic barrier. While for E. angustifolius, instead of a localised and sharp disjunction, we found a gradient of genetic differentiation across a much wider geographic area. Our findings suggest that biogeographic features may have different impacts on related species, and that generalisations on evolutionary patterns can be untenable without considering a range of factors.
As a result we investigated in more detail the factors most likely to have differentially affected the distribution of closely related species. We achieved this by quantifying patterns of genetic connectivity among 11 co-distributed elaeocarps across a biogeographic barrier, the Black Mountain Corridor (BMC) in the Australian Wet Tropics (AWT). We analysed a combination of allelic and flanking region sequence data from microsatellite markers, and evaluated the relative influence of environmental preferences and functional traits on genetic diversity and gene flow (Rossetto et al. 2009). The results indicate that only in three species geographic structuring of haplotype distribution reflects a north vs. south of the BMC pattern. Environmental factors linked with altitude were recognized as affecting genetic trends, but the selective processes operating on upland species appear to be associated with competitiveness and regeneration opportunities on poor soil types rather than climate variables alone. In contrast to previous observations within southeastern Australian rainforests, genetic differentiation in the AWT appears to be associated with small- rather than large-fruited species highlighting how external factors can influence the dispersal dimension. Overall, this study emphasizes the importance of considering functional and environmental factors when attempting generalisations on landscape-level patterns of genetic variation.

Main publications arising from this project
  • Crayn, DM, M Rossetto, DJ Maynard (2006) Molecular phylogeny and dating reveals an Oligo-Miocene radiation of dry-adapted shrubs (Tremandraceae) from rainforest tree progenitors (Elaeocarpaceae) in Australia. American Journal of Botany 93: 1168-1182.
  • Maynard, D, D Crayn, M Rossetto, R Kooyman, M Coode (2008) Elaeocarpus sedentarius sp. nov. (Elaeocarpaceae) – morphometric analysis of a new, rare species from eastern Australia. Australian Systematic Botany 21: 192-200.
  • Rossetto, M, D Crayn, A Ford, P Ridgeway, P Rymer (2007) The comparative study of range-wide genetic structure across related, co-distributed rainforest trees reveals contrasting evolutionary histories. Australian Journal of Botany 55(4): 416-424.
  • Rossetto, M, R Kooyman, W Sherwin, R Jones (2008) Dispersal limitations, rather than bottlenecks or habitat specificity, can restrict the distribution of rare and endemic rainforest trees. American Journal of Botany 95:321-329.
  • Rossetto, M, D Crayn, A Ford, R Mellick, K Sommerville (2009) The influence of environment and life-history traits on the distribution of genes and individuals: a comparative study of 11 rainforest trees. Molecular Ecology 18: 1422-1438.

Figure 1. Leaves of Elaeocarpus elliffii showing a diagnostic feature of prominent domatia.


Figure 2. Elaeocarpus angustifolius seedlings germinating after partial digestion of the fruit by a cassowary. Dispersal potential is an important factor explaining the distribution of these rainforest trees. 


Figure 3. Composite image showing Elaeocarpus bancroftii in flower, as well as variation in fruit and flower morphology among Wet Tropics species (the map identifies some of the collection sites used in this study).