Botanic Gardens Trust, Sydney, Australia

Proteaceae - shifting species boundaries

Maurizio Rossetto - Manager Evolutionary Ecology, Principal Research Scientist

The objective of this study on speciation in Proteaceae (funded by the Hermon Slade Foundation) as to obtain a better understanding of evolutionary processes in the Australian flora by investigating selected sister groups within this important family. The study genera (Telopea and Lomatia) were selected because relevant background knowledge and techniques were available from previous research, and because their distribution and diversification patterns were relevant to our target evolutionary questions: which reproductive barriers initiate speciation? Which barriers maintain reproductive isolation when species occur together (sympatry) and what are the selective forces that establish them? The story that is unraveling is compelling.

First, we found that temperature-dependent reproductive barriers (works done by Cathy Offord) between altitudinally separated populations of the NSW Waratah (Telopea speciosissima) were responsible for their morphological and genetic differentiation (Rossetto et al. 2011). Through follow up studies (in collaboration with Paul Rymer at University of Western Sydney) on gene expression and ecophysiology we are in the process of validating the adaptive nature of these differences (i.e. populations are better adapted to their local environmental conditions).

We also discovered that as climatic conditions change through time, existing reproductive barriers can be broken. In a study extending to all continental waratahs, we found that the repeated genetic signatures of hybridisation occurred across the whole genus (Rossetto et al. 2012). By using environmental niche modeling to investigate temporal changes in habitat availability (work done by Chris Allen) we showed that speciation was mostly driven by allopatric differentiation (i.e. geographic separation) followed by recurrent localised hybrid-isation events. Widespread gene flow among Telopea was unexpected, but it provided some novel insights on how locally adapted lineages might be able to respond to rapidly changing conditions.

In contrast, a genetic signal of hybridisation was expected across the genus Lomatia. What surprised us was the level at which hybridisation occurs. In her PhD study, Melita Milner (co-supervised by M. Rossetto, Peter Weston and M. Crip at the Australian National University) using chloroplast DNA (Milner et al 2012) and nuclear DNA (Milner et al in prep) showed that genetic diversity across five Lomatia species mostly represents geographic rather than taxonomic boundaries. Melita showed that within a confined region different species are genetically similar. She also demonstrated that haplotype-sharing is the result of a recurrent history of hybridisation in addition to incomplete lineage sorting (the sharing of ancestral alleles). Fine scale population genetic studies by Honours student Emma McIntosh (co-supervised by M. Rossetto, Peter Weston and G. Wardle at University of Sydney) showed that the degree of hybridisation varies at sites where species overlap according to local conditions and history. Interestingly, the genetic signature of past hybridisation events can still be recovered at sites that do not show morphological evidence of crosses between species (McIntosh et al. submitted). Emma’s findings suggest that despite gene flow, strong selective filters maintain ecotypical distinctions between species.

Overall we are discovering that gene flow between species creates a fascinating tension between maintaining strong adaptive differentiation and responding to rapidly changing environmental and ecological conditions. We hope to continue exploring this system in detail with targeted fine scale studies, and genomic analyses.






Distribution of nuclear DNA diversity (above) and chloroplast DNA (below) diversity among five Lomatia species. The figure shows that for both genomes (nuclear and plastid) diversity is distributed geographically rather than taxonomically (i.e. does not emulate the distribution of recognised species), suggesting high levels of hybridisation between these morphologically differentiated taxa (Milner et al. in prep).