CR2025_38 Shining a light on short-term response of insects and their predators to climate extremes: harnessing the potential of automated AI camera and audio systems

Lead Supervisor: David Roy, UK Centre for Ecology and Hydrology

Email: dbr@ceh.ac.uk

Co-supervisors: Tom Oliver, Department of Ecology and Evolutionary Biology, University of Reading; Jenna Lawson, UK Centre for Ecology and Hydrology; Mike Morecroft, Natural England

Declines in wildlife are increasingly being reported across the globe, giving stark warnings for the perilous state of biodiversity. Alarming reports of a 75% decline in insect biomass from Germany, have led the media to predict an ‘Insect Armageddon’ due to the vital role of insects in all ecosystems (e.g. as food for birds and mammals, recycling nutrients, pollinating crops) and as indicators of climate change impacts. Recent drought conditions across Europe have highlighted a research need to understand short-term (within and between days, weeks and months) responses to climate extremes in the context of longer-term trends.

Emerging camera and acoustic technologies offer huge potential for a step-change in the quality and quantity of biodiversity that can be collected to study short-term population dynamics. Automated sensors, deep learning, bioacoustics and computer vision are starting to deliver continuous, high temporal resolution and more standardised monitoring of insects, bats and birds.

This PhD project will use novel biodiversity monitoring data streams and existing environmental driver datasets to address the following questions: 1) How do insect activity patterns relate to weather conditions, particularly extremes in temperature (heatwaves and cold snaps); 2) How do predators (bats and insectivorous birds) respond to short-term fluctuations in food availability, particularly during critical periods such as when provisioning young; 3) How does habitat context mitigate the impacts of extreme events on insect biomass (e.g. do cooler habitat conditions buffer the impacts of extreme temperatures); 4) how do short-term population dynamics relate to long-term, large-scale fluctuations in insect numbers; 5) how are drought-sensitive insects predicted to respond to future climate and land-use scenario in the next 50 years.

This PhD studentship will capitalise on new cameras currently being deployed across the world in other 30 countries that span climate and land-use intensity gradients.  Combining with AI software developments for identifying species groups, and high resolution meteorological data will enable novel high-temporal-resolution research on the impacts of climate extremes on biodiversity.  To understand longer-term insect population dynamics and make future predictions, the studentship will request access to (i) datasets co-ordinated by DR under the European Butterfly Monitoring Scheme (eBMS), providing assessments of butterfly population from 12 regions of Europe and comparable butterfly monitoring schemes from at least 2 regions of North America; (ii) long-running assessments of terrestrial insect abundance for moths and aphids in the UK; (iii) data on bird and bat population dynamics from a NERC project (DRUID) and (iv) recently published Spatially explicit Projections of EnvironmEntal Drivers (SPEED).

New modelling approaches will be developed to maximise the potential of automated sensor data (cameras and audio), including the development of machine learning approaches to train image and audio classifiers and to develop algorithms to derive key biodiversity metrics (abundance, biomass, diversity etc).  Life history and methodological differences amongst existing insect monitoring approaches, in addition to differences in spatial and temporal resolutions, provide statistical challenges.  This PhD studentship will develop and apply approaches to predict biodiversity responses under future climate and land-use scenarios.

You will join a team of students and researchers focused on understanding the impacts of environment change (e.g. climate change, biological invasions, land-use) on biodiversity and ecosystems, with plenty of opportunities to learn new skills and benefit from mentoring from experts in their field.

This is a CASE studentship with Natural England, the national conservation agency for England, which has a large science team with complementary expertise in the practical application of monitoring methodology and world-leading scientific expertise in climate change adaptation.  You will have a supervisor from NE and access to its sites, offices and data as well as the opportunity for a placement.

Training opportunities: 

You will receive training in data science approaches that are revolutionising ecology research. In particularly, working with sensor data and the challenges of working with big data. Data science training will include high performance computing, AI and machine learning for image and bioacoustics data, together with Bayesian approaches for data analysis and model fitting. This will be combined with training in data analysis for ecological insight into the response of insects and their predators (birds and bats) to climate extremes. You will also learn how to translate your research into practical advice for climate change adaptation for nature conservation.

Student profile: 

We seek highly numerate candidates with an interest in novel technologies and developing skills in ecological modelling and conservation biology. Ideally, candidates will have an MSc with previous research project experience involving quantitative analysis, e.g. statistical ecology or process modelling. An additional skill is the ability to interact with a range of stakeholders, including data providers as well and NGOs and statutory agencies for which the project results are relevant. A desirable trait is a broad outlook on wider environmental issues to put these biodiversity change results in context and best realise impact from the research in this topical area. 

Please note: Due to the nature of this project and to comply with visa regulations, only Home students should apply.

Co-Sponsorship details: 

This project will receive a CASE award from Natural England.