Measuring the unmeasurable? Understanding uncertainty in biodiversity metrics

Project Description

Global efforts to save biodiversity revolve around measurable targets. They are the benchmark of both international and national policies, and the means by which we understand our progress (or not) towards conservation goals. And, as biodiversity becomes an increasing part of the agenda in the business and financial sectors, the drive towards quantifying biodiversity is only growing, whether to enable businesses to quantify their impacts, or for emerging financial mechanisms like biodiversity credits.

However, it is extremely difficult to quantify biodiversity, since it is neither a simple observable quantity, nor valuable in only one way. For this reason, there have been many recent proposals to quantify ecosystems by taking composite indices, which aggregate very different individual metrics. Most of these individual metrics are associated with high uncertainty bounds, which motivates the initial research questions of this project: How can these uncertainties be propagated through to the final composite indicator? How can uncertainties be better acknowledged? These questions could, for example, be tackled using a hierarchical Bayesian model which the student develops, though other statistical approaches are equally possible.

From this springboard, the student is free to take the project in a direction they choose. For example, how does uncertainty affect how we measure progress towards policy targets (such as England’s Environmental Targets, or the UN Mandated Global Biodiversity Framework)? How vulnerable are different choices of measurement uncertainty quantification to manipulation (for example to situations where a target is ‘met’, but outcomes on the ground don’t reflect this)?

The project will start with methodological investigations, but many opportunities exist for applied collaborations with organisations that work with biodiversity data, for example the UK Centre for Ecology and Hydrology or NatureScot.

Research questions

• How can estimates of uncertainty be combined for (linear and non-linear) combinations of biodiversity metrics?
• How does the treatment of uncertainty affect the performance of biodiversity metrics?
• How does choice of metric, and uncertainty treatment, affect conclusions about the state of biodiversity?

Methodology

Year 1: Conduct a literature review, familiarisation with different biodiversity measures, Bayesian statistics, and uncertainty quantification in large data settings; answer Q1 and start working on Q2.

Year 2: Answer Q2, submit a first manuscript about the results and findings. Start investigating possible avenues for Q3.

Year 3: Engage with policy stakeholders regarding Q3 and submit a manuscript discussing and, if possible, answering Q3. 

Training

A comprehensive training programme will be provided comprising both specialist scientific training and generic transferable and professional skills. The student will be encouraged to attend four one-week courses by the Academy for PhD Training in Statistics (APTS), and weekly `Biosphere' research group meetings to learn more about biodiversity research. There is a lot of potential to investigate professional placements at organisations involved in biodiversity metrics and measurement, e.g. the UK Centre for Ecology and Hydrology, or Nature Scot. Both the School of GeoSciences and the School of Mathematics at the University of Edinburgh have a large research student cohort that will provide peer-support throughout the research programme.

Requirements

We aim to recruit an enthusiastic student with a background in ecology, statistics, or a related discipline. As the project is deliberately open ended, we are seeking an independent student who is excited to explore their own ideas.

The student should have some familiarity with modern statistical methods as well as some knowledge of coding. Familiarity with the realities of ecological field work would be a big bonus.

How to Apply

See full project description here: https://e4-dtp.ed.ac.uk/e5-dtp/supervisor-led-projects/project?item=1719

And details on funding etc and how to apply here: https://e4-dtp.ed.ac.uk/e5-dtp

Note that as a NERC DTP this studentship has a somewhat convoluted application process: After the deadline, supervisors (Hannah and Torben) will select 1-2 applicants to put forward, these go into a pool alongside other student+project combinations from other supervisors. The DTP team then select, from the pool of student+project applicants, a subset to interview, and finally award offers to the best interviewees. This means not all projects go ahead.

If seriously interested in this project you are strongly encouraged to contact Hannah or Torben to discuss further.