Project Description

This project will examine the evolution of reproductive traits that are critical to the persistence of plant populations in heterogeneous environments, and apply that knowledge to understanding differential vulnerability of tropical tree species to loss of genetic diversity following population fragmentation. 


Plants are dependent on movement of pollen and seeds for sustaining gene flow within and between populations, which is critical to maintenance of genetic and species diversity. However the dispersal of pollen and seeds across landscapes may be disrupted by anthropogenic disturbances, such as logging, that change the density of reproductive individuals, or create barriers to movement through habitat fragmentation. These disturbances may be particularly severe for many tropical tree populations, which have been exposed to rapid transitions in their population density and habitat structure as a consequence of recent deforestation and fragmentation. These transitions have been particularly severe in Southeast Asia, which are dominated by trees in the Dipterocarpaceae. Dipterocarps represent a high-value source of tropical timbers and logging based on extraction of these trees underpins local economies and livelihoods across the region.  

Dipterocarps possess a suite of reproductive traits that induces enhanced vulnerability to loss of genetic diversity in response to reductions in population size or habitat fragmentation. They have evolved a ‘general flowering – mast fruiting’ reproductive phenology, resulting in infrequent and episodic recruitment following community-wide flowering events. They are pollinated by a wide variety of insects ranging from thrips to giant honeybees, and there is evidence that competition for pollinator services limits pollination success during these flowering events, particularly for species that produce a large number of small flowers. Importantly, pollinator body size is linked to their mobility, so mean pollen dispersal distance declines for small-flowered species pollinated by small insects. Short-distance pollen dispersal confers a greater likelihood of sib-sib mating in plants, particularly for those with inherently limited seed dispersal such as dipterocarps. Such species are characterised by a high degree of fine-scale genetic structuring within populations, and an inherent risk of inbreeding costs in response to limited pollinator mobility.  

These attributes of their reproductive biology may help explain a syndrome of trade-offs linking flower size, pollination success, pollinator mobility and spatial genetic structure across multiple dipterocarp species. These attributes are shared by many other groups of angiosperms, and it is possible that similar syndromes may exist more generally in many insect-pollinated plant groups. However, this syndrome linking plant traits to population genetic structure has never been framed or tested theoretically, and its implications for the actions required to conserve different plants have not been explored. 


Competition among coexisting species for pollination services drives reciprocal evolution of pollinator and plant reproductive traits, including flower size, pollinator size and pollinator mobility, leading to 

  • A negative trade-off between flower size and flower number across species 
  • A positive association between flower size and pollinator body size 
  • A positive association between flower size and pollination success 
  • A positive association between flower size and pollen dispersal distance 
  • A negative association between flower size and frequency of sib-sib mating 
  • A negative association between flower size and fine-scale population genetic structure 
  • Equalisation of reproductive success 

Flower traits predict vulnerability to loss of genetic diversity following reduction in population size or habitat fragmentation, e.g. 

  • Failure of cross-pollination for small-flowered species with small pollinators, leading to reduced fruit crop size and recruitment failure
  • Loss of heterozygosity among progeny in habitat fragments, particularly those with small flowers and higher fine-scale spatial genetic structure

Essential and desirable skills: 

Essential: applicants should possess a strong background in quantitative ecology.  

Desirable: with an interest in developing dynamic eco-evolutionary models and/or bioinformatics approaches to developing the project. 


David Burslem

Primary Supervisor:

Profile: David Burslem
Institution: University of Aberdeen
Department/School: School of Biological Sciences

Paul Caplat

Secondary Supervisor:

Profile: Paul Caplat
Institution: Queen's University, Belfast
Department/School: School of Biological Sciences

Justin Travis

Additional Supervisor:

Profile: Justin Travis
Institution: University of Aberdeen
Department/School: School of Biological Sciences

Greta Bocedi

Additional Supervisor:

Profile: Greta Bocedi
Institution: University of Aberdeen
Department/School: School of Biological Sciences

Natalie Pilakouta

Additional Supervisor:

Profile: Natalie Pilakouta
Institution: University of Aberdeen
Department/School: School of Biological Sciences

Additional Supervisor:

External supervisors in CASE partner: 

  1. Dr Chris Kettle ( Alliance-CIAT, Rome, Italy 
  1. Dr Riina Jalonen ( Alliance-CIAT, Kuala Lumpur, Malaysia 


Bocedi, G., et al. (2021) Ecography, 44, 1453-1462. doi:10.1111/ecog.05687 

Kettle, C.J., et al. (2011) PLoS ONE, 6: e16111. doi:10.1371/journal.pone.0016111

Nicholson, C.C., et al. (2019) Journal of Applied Ecology, 56, 618–628. doi: 10.1111/1365-2664.13333

Research Methods

The final combination of methods will depend on the interests and background of the student appointed. The following opportunities are available through the research team. 

 Field sampling to estimate gene flow via pollen and seed dispersal 

Working with existing mapped populations of adult trees of multiple dipterocarp species in Sabah, Malaysia, we will sample and genotype mother trees, potential pollen donors and progeny (seeds or seedlings depending on availability) for continuous populations in intact habitat and fragmented populations in human-modified landscapes dominated by oil palm plantations. The species will be selected to sample a suitable range of flower size within the family, in order to capture effects due to differences in pollinator body size and mobility.  

Bioinformatic and data analyses 

Samples will be sequenced and bioinformatic methods will be used to determine the mean and variance of pollen and seed dispersal distances within and between populations and species. The data will also be suitable for estimating fine-scale spatial genetic structure and coefficients of inbreeding among adults and progeny across species and populations. These analyses will be conducted in conjunction with partners at Bioversity International in Rome and Malaysia to allow integration with their existing international research programmes on conservation genetics of tropical trees. 

 Eco-evolutionary modelling 

The RangeShifter platform will be adapted to simulate the characteristics of out-crossing monecious plants with short-distance seed dispersal. The aims of these simulations will be to test the hypothesis that competition for pollinator services during a general flowering event is sufficient to stabilize species coexistence, and to examine consequences of habitat fragmentation for genetic diversity, fecundity and persistence of multiple dipterocarp species with varying reproductive traits. Parameterization of RangeShifter will be achieved using existing demographic and dispersal data for dipterocarp species from the region, and the data newly collected by the student. 

Expected Training Provision

The training for this PhD studentship has four main components:

Sampling design and fieldwork  

Training will be provided in the theoretical principles and practical techniques required for field sampling of plant material for estimating gene flow and metrics of population genetic structure. This will require balancing considerations such as sample size, quantity, accessibility of sampling locations and cost of analysis, and engagement with logistical issues such as research access, export and import permits. 

Molecular and bioinformatic analyses 

The student will obtain training in all the laboratory protocols required for the molecular and bioinformatic components of this project through cohort training and courses offered through the University of Aberdeen’s Centre for Genome Enabled Biology and Medicine (CGEBM), as well as guidance from supervisors and group members. The student will also benefit from close interaction with external co-supervisor Dr Chris Kettle, who leads Bioversity International’s cross-cutting research team working on conservation and sustainable use of socio–economically and ecologically important trees and their genetic diversity. Dr Kettle is also Group Leader in Applied Molecular Ecology, in the Department of Environmental System Science, ETH Zurich, Switzerland, and has significant expertise in the conservation genetics of tropical trees. 

Eco-evolutionary modelling 

The student will be co-located within a large research group of ecological and evolutionary modellers where the RangeShifter platform was developed. Training is use of RangeShifter and other modelling approaches is delivered routinely via formal and informal settings by group members, including courses held in Aberdeen, online and internationally. The group is currently collaborating with external partners in Germany, who developed RangeShiftR (Malchow et al. 2021), which provides multiple opportunities to interface the specific functionality of the RangeShifter platform with the power of R software environment.

Complementary training in transferable skills 

Training in core scientific skills will be provided through the Quadrat DTP, including presentation skills, time management, team-working, communication skills and paper writing. Since this studentship involves close working with an external international organisation embedded within UN-FAO (Bioversity International), there will be opportunities for the student to gain direct experience of translating research into developmental impact. This will be achieved during an internship with Bioversity International during the latter stages of the PhD. Training will be provided to ensure that the PhD student disseminates the research via research and review papers in peer-reviewed journals and yearly talks or posters at conferences (e.g. annual meetings of the British Ecological Society, Irish Ecological Association, Ecological Society of America, Oikos society), as well as outlets targeting a non-specialist audience (e.g. Conversation) and social media. 


The project will contribute new insights into the mechanisms that generate and maintain diversity within and between tropical tree species. This is important because loss or degradation of tropical tree populations undermine efforts to protect biodiversity and sustain the ecosystem services they deliver. Multiple national and international policies and legal instruments recognise the importance of maintaining and/or restoring tree populations. One of the key drivers of tropical forest degradation is the failure of many tree species to reproduce successfully in human-modified environments, particularly where anthropogenic disturbance has reduced the density of reproductive adults and separated them in space through habitat fragmentation. Pervasive recruitment failure may arise because animal-mediated pollination becomes disrupted by the over-dispersion of adult trees or reductions in pollinator abundance. This project will provide insights into which traits confer differential vulnerability to cross-pollination among tropical trees, and therefore help to prioritize conservation efforts. It will also identify landscape configurations that might inhibit out-crossing for multiple taxa with different traits, enabling a deeper understanding of how landscape structure interacts with individual species’ traits to determine community assembly in habitat fragments. 

This information is important to the design of restoration strategies for degraded and fragmented forests globally. Approaches involving tree planting require a large supply of genetically diverse and well-adapted seedlings, but planting material sourced from residual tree populations in forest fragments may instead be inbred and lacking in vigour. Similarly, reliance on natural regeneration will only be successful if new seedlings of the desired species are recruited in sufficient numbers, otherwise these is a risk that sites will become dominated by pioneer trees and lianas. Greater knowledge of the traits that confer successful out-crossing in fragmented populations, and the drivers of loss of genetic diversity, will help inform decision making over the distribution of restoration effort at landscape scales.  

Proposed Supervision

Supervision for this student will be provided by David Burslem (University of Aberdeen) and Paul Caplat (Queen’s University Belfast), with additional contributions from Justin Travis, Greta Bocedi and Natalie Pilakouta (University of Aberdeen), and external supervision provided by Chris Kettle and Riina Jalonen (Bioversity International). David Burslem is a plant ecologist with a long history of working on the mechanisms that maintain the genetic and species diversity of tropical forest trees, including aspects of their spatial ecology, pollination biology and population genetics relevant to this proposal. He has contacts with local organisations and knowledge of field sites required to facilitate the fieldwork, and access to current data-sets for model fitting in the unlikely event that international travel is not possible throughout the duration of the studentship. Paul Caplat is a landscape and population ecologist with particular expertise in modelling spatial processes including pollen and seed dispersal. Justin Travis is a theoretical ecologist working on spatial eco-evolutionary dynamics who will support model development using the RangeShifter model, which was developed within his group and has recently been translated onto the R platform. He also has current research and contacts within Southeast Asia. Greta Bocedi works on the ecological, evolutionary and eco-evolutionary dynamics of species adaptations and behaviours in space and during environmental changes, their consequences for population dynamics, persistence and diversification, and species responses to rapid environmental changes. Natalie Pilakouta is an evolutionary biologist interested in how environmental change alters organismal behaviour, integrating perspectives from behavioural ecology, evolutionary biology, ecophysiology and developmental biology. David, Justin and Paul have a history of successful collaboration on recent NERC-funded research. 

Proposed Timetable

The timetable for the research would be developed by discussion between the supervisors and the successful candidate during the early stages of the project, to ensure that objectives and milestones are realistic whilst also reflecting the aspirations of the candidate in terms of focus and training requirements. Therefore, the following timetable should be regarded as indicative only. 

Year 1 

The first six months will be devoted to induction, training and development of a research proposal, which will include exploration of appropriate modelling platforms. This phase will include intensive reading, preparation of a literature review and attending compulsory generic training modules. A final task will be preparation for fieldwork, which will involve submission of all necessary permit applications, import licenses and risk assessments. 

The second six months would be an opportunity to conduct the first field season for sample collection, and completion of the nine-month assessment progression requirement. 

Year 2 

The first six months will focus on molecular analysis of samples collected in Year 1, followed by bioinformatic analyses of the data. Where necessary, these activities will be preceded by training. Presentation of a poster at the Annual Meeting of the British Ecological Society would be a reasonable target. 

The second six months will focus on further model development and writing a paper based on the preceding work, followed by a second field season to fill gaps in sample coverage. 

Year 3  

The main task for the third year will be to complete all molecular and bioinformatics analyses, implement model runs and then initiate thesis preparation. An internship with Bioversity International would be possible during this period, especially in the context of outreach and dissemination activities. Oral presentation at a conference would be expected. 

Year 4 (six months) 

The entire focus will be on completing the thesis chapters and associated papers.  


  • biodiversity
  • environmental-management


CASE Partnership with Bioversity International (Alliance-CIAT) 

This project will be conducted in partnership with co-supervisors on the staff of Bioversity International, which is an international organisation headquartered in Rome and with offices in many tropical countries including Malaysia.  The purpose of Bioversity International is to ‘deliver research-based solutions that harness agricultural biodiversity and sustainably transform food systems to improve people’s lives in a climate crisis.’ The specific partners in this project and co-supervisors of this project are Dr Chris Kettle, an applied molecular ecologist based in Rome, and Dr Riina Jalonen who works on the conservation of tropical trees with a base in Malaysia. The student will benefit from this linkage by participating in an internship at Bioversity International during the studentship. 

View All Projects