Harnessing AI Tools and DNA-based Monitoring to Enhance Arthropod Biodiversity Assessments

Arthropods are vital for forest ecosystems, but their populations are being threatened by habitat disturbances, climate change and invasive species. This project uses AI and DNA metabarcoding to improve biodiversity assessments in Northern Ireland’s commercial forests. The student will analyse samples, develop identification tools, and receive training in AI, molecular techniques, and data management, preparing for careers in data science and ecology.

Scientific Background:

Arthropods play critical roles in forest ecosystems, from pollination to decomposition. In commercial forests like spruce, pine, and larch, they influence ecological balance and productivity. However, habitat disturbances, climate change, and invasive species are altering arthropod populations, threatening ecosystem stability. Traditional monitoring methods are labour-intensive and unsuitable for large-scale use. This project employs AI and DNA metabarcoding to transform arthropod biodiversity assessments, enhancing efficiency and accuracy. By integrating molecular and computational tools, it aims to assess the health and sustainability of Northern Ireland’s commercial forests.

Research Methodology:

The student will analyse entomological samples collected over three years from Northern Ireland’s commercial forests:

• Image Processing and Identification: Using neural networks trained on public biodiversity databases and curated AFBI records and images to classify arthropod species from high-resolution images.
• Tool Development: Designing a user-friendly tool that combines predictive modelling with image-based species identification for researchers and forest managers.
• Validation: Confirming AI-based identifications through DNA metabarcoding and traditional taxonomy to ensure accuracy.

Training:

The student will receive hands-on training in AI programming and molecular techniques, primarily based at Queen’s University Belfast, with organized visits to the University of Strathclyde. They will develop expertise in programming, machine learning, neural networks, DNA metabarcoding, and large dataset management. Training will also cover image analysis and integrating diverse data types. Due to the highly computational nature of this project, this opportunity would be best suited to candidates with proven previous experience in programming (Python, R). An 18-month internship with AFBI project partners will provide experience in entomology, taxonomy, and policy-relevant research. This interdisciplinary training will prepare the student for careers in data science, ecological research, and environmental management.