Mapping LIFE on Earth

Human-driven habitat loss is recognised as the greatest cause of biodiversity loss, but we lack robust, spatially explicit metrics quantifying the impacts of anthropogenic changes in habitat extent on species' extinctions. LIFE is our new metric that uses a persistence score approach that combines ecologies and land-cover data whilst considering the cumulative non-linear impact of past habitat loss on species' probability of extinction. We apply large-scale computing to map ~30k species of terrestrial vertebrates and provide quantitative estimates of the marginal changes in the expected number of extinctions caused by converting remaining natural vegetation to agriculture, and also by restoring farmland to natural habitat. We are also investigating many of the conservation opportunities opened up via its estimates of the impact on extinctions of diverse actions that change land cover, from individual dietary choices through to global protected area development.

LIFE v1

Our efforts through 2023-24 were focussed on building the first version of the LIFE metric and addressing peer review contents, with the research expertly lead by Alison Eyres. The LIFE: A metric for mapping the impact of land-cover change on global extinctions paper appeared in publication at the Royal Society in early 2025, and even covered by Mongabay.

The computational challenges of generating global maps for ~30k species at 1 arc-minute resolution required significant high-performance computing resources and careful attention to dataset versioning and reproducibility. As Patrick Ferris notes, large ecological datasets are inherently difficult to version and reproduce, making our approach of coupling persistence scores with HPC infrastructure particularly important for ensuring scientific reproducibility.

Our research efforts in 2025 are focussed on improving the resolution of the persistence maps, increasing the coverage of species, and performing more analyses to identify newer conservation opportunities. This work is part of broader planetary computing efforts to make global-scale biodiversity data accessible for decision-making.

Computational Challenges and Infrastructure

The LIFE metric represents a huge computational ecology challenge, requiring processing of species occurrence data, habitat maps, and persistence calculations across multiple spatial and temporal scales. Our work highlights the broader challenges in computational conservation, where the scale of ecological data now requires sophisticated computational infrastructure.

Key technical challenges we've addressed include Managing and versioning terabyte-scale biodiversity datasets across time, scaling persistence score calculations across 30,000+ species, ensuring reproducible computational workflows for ecological modeling, and balancing computational efficiency with ecological model accuracy. This computational ecology approach is increasingly vital as conservation decisions require rapid, evidence-based analysis of large-scale environmental data, and I am hosting the 2nd outing of Programming for the Planet in October 2025 to continue this conversation.

Mapping other spatial threats

Unfortunately, individual species are affected by anthropogenic threats beyond simply habitat loss from landuse change, including hunting, agricultural practices and the introduction of invasive species. Emilio Luz-Ricca is conducting his PhD research on this topic, in particular focussing on per-species abundances and threats. Charles Emogor -- who completed his PhD in 2024 on threats to pangolins via hunting -- is also joining the Computer Lab as a Schmidt Sciences fellow and applying machine learning to predicting sources of hunting pressures on wild species.

Tying anthropogenic activity such as food consumption to biodiversity.

Agriculturally-driven habitat degradation and destruction is the biggest threat to global biodiversity, and so an exciting line of work that Thomas Ball has been leading is to tie the LIFE metric with food consumption and production data and provenance modelling in order to figure out the impact of what we eat on species extinctions. The FOOD metric papers show that despite marked differences in per-capita impacts across countries, there are consistent patterns that could be leveraged for mitigating harm to biodiversity.

This work connects to broader questions about sustainable food systems and how computational tools can help consumers and policymakers understand the biodiversity consequences of dietary and agricultural choices. We're continuing to work on refining this data and analysis, particularly via higher resolution supply chain datasets and crop yield data.