IAP-24-069
Origin and evolutionary dynamics of rainforest assembly in the Fiji Islands
Background
Rainforests are among Earth’s most species-rich ecosystems, yet the processes driving their assembly and evolution remain hotly debated [1-3]. Continental rainforests, such as those in South America or Southeast Asia, have experienced complex and lengthy geological histories, marked by significant species turnover and only sparse fossil records due to often poor fossilization [1,4]. These challenges have limited our understanding of when and how rainforest biodiversity assembled, particularly the interplay between ancient lineages and recent speciation.
The Fijian archipelago, a global biodiversity hotspot in the South Pacific [5], presents a unique system to investigate these fundamental questions. Unlike continental rainforests, the Fiji Islands have a relatively recent and well-understood geological history and a sizeable plant diversity (1,643 native vascular plant species, 51% endemic) making them an ideal, scalable system for rainforest assembly. These islands’ variation in area, age, elevation, and isolation provides natural replicates to study biogeographic history, and diversification dynamics in rainforest ecosystems.
This project addresses a critical gap in rainforest evolution by investigating the timing, geographic origins, and ecological drivers of diversification and endemism in Fiji’s flora. Fiji offers a rare, scalable system to study these processes in detail, enabling the development of an ecosystem assembly model that is potentially transferable to larger, more complex rainforests.
The project will address this major knowledge gap following three main objectives:
Objective 1: Determining the global biotic interchange of the Fijian flora. Using a new comprehensive global plant phylogenies including most or all vascular plant species, you will conduct large-scale biogeographic analyses to trace the origins, timing, and frequency of immigration events to and from Fiji and within its archipelago, focusing on the largest islands: Viti Levu, Vanua Levu, and Taveuni. This will clarify whether Fijian rainforest assembly relied on ancient continental immigrants or is driven primarily by recent long-distance dispersal and in-situ speciation. Additionally, it will allow comparisons of rainforest assembly across islands with distinct geological ages—Viti Levu (30 Myrs), Vanua Levu (4 Myrs), and Taveuni (0.8 Myrs)—and unveil the relationships between rainforest and other Fijian biomes, such as montane and dry forests.
Objective 2: Evolutionary dynamics of the most speciose endemic Fijian lineages. You will then zoom on a few lineages with the highest numbers of endemic species including Psychotria (Rubiaceae) with 79 indigenous species (72 endemic), Cyrtandra (Gesneriaceae) with 41 indigenous species (23 endemic), and Elaeocarpus (Elaeocarpaceae) with 23 indigenous species (21 endemic). You will generate phylogenomic trees to provide solid evolutionary frameworks to analyse their biogeography and diversification dynamics, relying on herbarium material but also on your own fieldwork in Fiji to sample additional species. You will then use state-of-the-art methods to infer biogeographic history and diversification dynamics. This fine-scale approach will provide a detailed spatio-temporal framework for the origin of key constituent lineages of Fijian rainforests.
Objective 3: Characterizing the spatial patterns in the evolutionary assembly of the Fijian floras. Using data from Objectives 1 and 2, you will develop spatial models of species richness and endemism across Fiji, integrating GIS data from GBIF, herbarium records, and field surveys. These models will be paired with a functional trait database (e.g., seed dispersal mechanisms, leaf size) to test the geographic, ecological, and evolutionary drivers of species richness. This approach will allow us to determine whether regions of high species richness coincide with areas of high lineage diversity, old lineages, or a mix of recent immigrants and endemics, offering insights into the processes of ecosystem assembly. It will also identify drivers of endemism, with substantial consequences for conservation.
Altogether, this PhD project will provide an unprecedented advance into rainforest evolution, and in turn provide substantial impact to plant conservation in Fiji.
Click on an image to expand
Image Captions
Fijian rainforest, Lavena, Taveuni island, photo G. Chomicki
Methodology
For Objective 1, you will use a new global plant phylogeny (updated from [6]), and BioGeoBEARS [7] and recent methods for ancestral range and range shift inference in mega-phylogenies [3,8].
For Objective 2, fieldwork will be conducted in Fiji to collect samples in addition to samples from herbaria (mostly Fiji [SUVA] and Kew [K] herbaria). You will use whole genome sequencing (Illumina) to generate DNA data. You will use concatenated likelihood analyses (e.g. IQ-TREE2 [9]), coalescence methods (ASTRAL-III [10]) and IQ-TREE-POMO [11] to generate solid phylogenetic trees, which will be time-calibrated using BEAST2 [12] and RevBayes [13]. Historical biogeography will be reconstructed using BioGeoBEARS, and diversification rates will be estimated using tools such as BAMM [14].
For Objective 3, you will generate high quality spatial data combining three sources: 1) publicly available data from gbif (https://www.gbif.org), 2) herbarium data, either digitized from JSTOR global plants (https://plants.jstor.org), and non-digitized, by visiting herbaria notably Kew (K) and Fiji (SUVA), and 3) compiling data from quantitative field surveys across Fiji, from collaborator M.T., which will enable you to generate maps of species richness and endemism. You will then generate a database of plant functional traits (notably seed dispersal, leaf size etc) to test for potential drivers of species richness and endemicity. Spatially explicit general linear models will be used to test the distinct drivers of species richness and endemicity.
Project Timeline
Year 1
For Objective 1, all biogeographic analyses will be carried out. For objective 2, all samples will be acquired from herbarium material (many already at hands), and fieldwork in Fiji will be carried out. DNA will be extracted for all samples, and they will be sequenced.
Year 2
All analyses from Objective 1 will be wrapped and written up for a thesis chapter and related publication. All phylogenomic analyses, molecular clock dating and diversification analyses from Objective 2 will be carried out. Start to compile the GIS data for Objective 3.
Year 3
All analyses from Objective 2 will be written up for a thesis chapter and related publication. Finish to compile GIS data for Objective 3, and carry out spatial analyses.
Year 3.5
All analyses from Objective 3 will be wrapped and written up for a thesis chapter and related publication. Thesis and publication writing.
Training
& Skills
The supervisory team combines a synergy of expertise across evolutionary biology, ecology, bioinformatics and plant systematics. You will receive an excellent training in a wide range of bioinformatics, phylogenetic comparative methods and molecular clock dating from G.C. and C.K., spatial analyses from P.S. as well as use of Durham supercomputer Hamilton. You will also be trained in field-based botany and plant systematics by G.C. and M.T. Moreover, you will benefit from the stimulating environment of the Chomicki lab, which includes 10 current members and a wide range of skills and research interests.
References & further reading
1. Morley, R.J. (2000). John Wiley & Sons.
2. Eiserhardt W.L. et al. (2017). New Phytol. 214:1408-1422.
3. Pérez-Escobar O.A et al. (2022). Trends Plant Sci. 27:364-378.
4. Pennington RT et al. (2004). Phil. Trans. Roy.Soc. B 359:1455–1464.
5. Myers, N. et al. (2000). Nature 403:853-858.
6. Smith, S.A., & Brown, J.W. (2018). Am. J. Bot.105: 302-314.
7. Matzke N.J. (2014). Syst. Biol. 63:951-970.
8. Antonelli A., et al. (2018). PNAS 115:6034-6039.
9. Minh B.Q. et al. (2020). Mo. Biol. Evol. 37:1530-1534.
10. Zhang, C. et al. (2018). BMC Bioinformatics 19:15-30.
11. Schrempf, D. et al. (2019). Mol. Biol. Evol. 36:1294-1301.
12. Bouckaert R. et al. (2014). PLoS Comput. Biol.10: e1003537.
13. Höhna S et al. (2016). Syst. Biol. 65:726-736.
14. Rabosky D.L. et al. (2014). Methods Ecol. Evol. 5:701-707.