Understanding ecosystem-wide responses to anthropogenic pressure

Lead Research Organisation: University College London
Department Name: Genetics Evolution and Environment

Abstract

Human-induced environmental pressures are altering ecological communities in ways not fully understood. In the face of a biodiversity loss crisis, understanding the impact of these pressures is crucial for future conservation efforts. In this thesis, I investigate wildlife responses to anthropogenic pressure and consider how biodiversity monitoring methods could be optimised. I use macro-ecological datasets and a field study in Nepal to investigate different methods used for biodiversity monitoring at varying spatial and ecological scales. I utilise cutting-edge machine learning technology to create an efficient pipeline for analysing camera trap data, providing metrics on behaviour change and community composition in response to anthropogenic pressure. In addition, novel automated methods for analysing camera trap data without species identification are proposed and explored. The findings presented in this thesis address the paradoxical needs of global biodiversity monitoring, where comprehensive global metrics must simultaneously account for the nuances of species' responses to different pressures.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
NE/S007229/1 30/09/2019 29/09/2028
2235776 Studentship NE/S007229/1 30/09/2019 29/06/2024 Peggy Bevan
NE/W502716/1 31/03/2021 30/03/2022
2235776 Studentship NE/W502716/1 30/09/2019 29/06/2024 Peggy Bevan
 
Description During this award I have performed research on the way we monitor biodiversity. Biodiversity is a broad term and therefore there are many ways to measure it, but there should be an international agreement on how biodiversity is assessed in order to move conservation and envrionmental protection forward. I have recently submitted for peer-review a macro-ecological study that looks at how biodiversity responds to human pressure at different spatial scales using species richness and total abundance as my metrics. This helps us to understand how global reporting on biodiversity trends is influenced by scale.
Exploitation Route These findings can contribute to discussions by major international bodies such as the Convention of Biological Diversity, to better understand how to set targets and measure progress towards these targets.
Sectors Environment

 
Description Through the use of cutting-edge machine learning technology I am researching new methods for monitoring biodiversity using camera traps and passive acoustic monitoring. This research is ongoing but I hope it will provide impactful tools that can be used by industry, the public sector and other researchers in the environment sector.
First Year Of Impact 2022
Sector Environment
Impact Types Policy & public services

 
Description Computer Vision for Ecology Summer School 
Organisation California Institute of Technology
Country United States 
Sector Academic/University 
PI Contribution I joined a summer school hosted by the california institute of technology where I received training and worked on a project that is part of my thesis. My contribution to make an outcome of this is to create a publication and include my tutors as co-authors. I also wrote a report outlining my work at the school for the organisers to feedback to funders.
Collaborator Contribution The summer school provided me with approx 2000USD of computing credits with Microsoft Azure for me to run machine learning models on.
Impact I was invited to speak about our work at the 'AI for the Natural Sciences' symposium hosted by the Natural History Museum as well as an AI for Ecology online workshop.
Start Year 2022