Leverhulme ‘Space for Nature’ Doctoral Scholars (PhD or MSc by Research+PhD) Bumblebee populations in the UK

Leveraging Machine Learning to Assess the Impact of Other Effective Area-Based Conservation Measures (OECMs) on Bumblebee Populations in the UK

The MSc by Research (if relevant) and PhD scholarships include a stipend (equivalent to the Research Councils UK National Minimum Doctoral Stipend; the 2024/25 rate is £19,237, which is not taxed income). Tuition fees are covered at the home student rate. The PhD scholarship comes with a £10,000 research and training fund.

The application deadline for this scholarship has passed. 

This award is open to home students only. To be classed as a home student, candidates must meet the following criteria and the associated residency requirements:

• Be a UK national or,
• Have settled status or,
• Have pre-settled status or,
• Have indefinite leave to remain or enter

Criteria

• Hold a 1 or 2.1 Bachelor's degree or, if applicable, a taught Master's degree at merit or distinction or MSc by Research. Please be aware that securing a PhD scholarship tends to be a competitive process, with most applicants holding a Master’s degree. We are therefore committed to trying to remove this barrier by offering fully funded MSc studentship at the University of Kent, particularly targeted towards individuals from lower-income or ethnic minority backgrounds. This project has access to a MSc by Research scholarship.
• Provide a CV. On the CV, please list the degree modules you have studied and provide the grade you were awarded for each one. Please also provide the overall grade you were awarded for your degree(s).
• Provide a covering letter, which outlines why you are interested in the PhD, no more than two A4 pages long. Any statement exceeding this limit will not be accepted.
• Complete our equality, diversity and inclusivity questionnaire in full via this link: https://universityofkent.qualtrics.com/jfe/form/SV_1zXpdlSFDOmWmDc
• Provide academic references in support of your application; these will be requested if you are successful through the shortlisting process.
• Shortlisted candidates will be interviewed by members of the supervisory team.
• Be able to start the MSc by Research or PhD programme in Sept 2024.

Leveraging Machine Learning to Assess the Impact of Other Effective Area-Based Conservation Measures (OECMs) on Bumblebee Populations in the UK

For more background information about the Leverhulme ‘Space for Nature’ Doctoral Scholars, please go to: www.kent.ac.uk/durrell-institute-conservation-ecology/research/leverhulme-scholars.

School of PhD registration: School of Mathematics, Statistics and Actuarial Science

PhD degree award: Statistics

Primary supervisor name: Dr Eleni Matechou (School of Mathematics, Statistics and Actuarial Science, UoK) (DICE, UoK)

Email address: e.matechou@kent.ac.uk

Co-supervisor name: Dr Christos Efstratiou (School of Computing, UoK)

Co-supervisor name: Dr Richard Comont (Science Manager, Bumblebee Conservation Trust)

Project Details: Monitoring bumblebee populations is imperative due to their critical role as pollinators. ‘Other Effective Area-Based Conservation Measures’ (OECMs) are places that are outside the protected site network, but which are managed in a nature-friendly way alongside management for other reasons. The OECM approach - conservation beyond nature reserves – has the potential to support bumblebees at a population scale for the first time. The extent of this support (foraging, nesting and overwintering resources) is highly dependent on the habitats present within the OECM and OECM location, size, shape, landscape connectivity, and surrounding habitats. However, no metrics or methods currently exist for quantifying the impact of OECM characteristics on bumblebee populations, understanding the potential value of OECMs for supporting these important pollinators and informing policy and conservation actions accordingly.

BBCT manage the only GB bumblebee monitoring scheme (BeeWalk), numbering approximately 800 volunteers, collecting monthly bumblebee count data (48,000 records in 2023). Additionally, opportunistic bumblebee sightings are recorded nationwide using mobile-phone apps (e.g. iRecord, iNaturalist). Together, these schemes provide unprecedented spatio-temporal data, alongside regularly-updated remote-sensed data associated with the locations and times of observations, including novel metrics (e.g. non-visible-spectrum reflectance).

This project brings together statistical modelling of the data-generating process with machine learning, including deep learning, techniques, to model and predict bumblebee populations comparing potential OECMs and existing protected areas in terms of their effect on bumblebee populations, using all available, diverse, data. The ongoing collection of bumblebee and remote-sensed data has the potential to provide near-real-time reliable feedback on the value of OECMs for bumblebee populations, and by extension the wider pollinator guild, informing conservation management as well as the concomitant benefits of pollination for surrounding land users.

This interdisciplinary project merges applied ecology, statistics, and computing to construct a robust modelling framework to infer bumblebee population dynamics and trends, link these to OECMs and their characteristics, and ultimately derive insights into the potential effectiveness of OECMs in bumblebee conservation and inform corresponding conservation strategies.

The selected candidate will receive comprehensive training in the latest machine learning and deep learning techniques. Suitable applicants have a strong background in statistics, mathematics or computing.