New tools will combine deep understanding of animal behaviour with AI to provide early warning systems for conservation


A photo collage showing from left to right: A Damselfish, Coral Field and Dr Sally Keith in a wetsuit wearing diving equipment © Umeed Mistry (Damselfish image), and Dr Sally Keith
Left to right: A Damselfish, Coral Field and Dr Sally Keith

Marine scientists are developing new tools to recognise changes in fish behaviour that indicate an ecosystem is under stress.

These new tools will provide an early warning system that will help conservationists recognise issues before they escalate, enabling earlier, and potentially less costly, intervention.

The system is only possible by harnessing the power of AI-based tools at multiple stages, from extracting behavioural data from video footage, to creating decision-support tools with conservation managers. Tools will be developed and deployed on coral reefs before transfer to UK waters.

“Global biodiversity loss is accelerating,” said Dr Sally Keith, Senior Lecturer and Turing AI Pioneer Fellow at Lancaster University. “Despite this urgency, conservation is often reactive, with interventions triggered only after individuals start to die. There is a pressing need for earlier detection of ecological risk, enabling more rapid conservation responses, before irreversible damage sets in.”

The ‘BEACON: Behavioural Ecology AI for CONservation action’ project led by Dr Keith will develop conservation tools that combine deep ecological understanding through an ambitious ecological field programme with cutting edge AI to understand when, and why, animal behaviour changes in response to disturbance.

The project is funded through Dr Keith’s three-year Turing AI Pioneer Interdisciplinary Fellowship of nearly £2.2 million by UK Research and Innovation (UKRI). These fellowships aim to give established researchers the time and resources needed to bring AI into their research programmes. Once recruited, the full interdisciplinary team spans 4 Senior Research Associates and a specialist engineer, bringing together expertise in AI, behavioural ecology, conservation decision-making and edge AI devices.

Because animal behaviour often responds rapidly to disturbances, these behavioural changes can provide subtle signals of stress.

However, monitoring animal behaviour, especially in marine environments, is extremely time-consuming, making it unfeasible to do at scale. This constraint has limited its usefulness for conservation.

BEACON aims to overcome these hurdles by developing AI tools that enable the gathering and interpretation of animal behavioural data across large scales. These tools will automate the collection, analysis and interpretation of behavioural data in the field, before flagging unusual behaviours that indicate stress, and recommending actions for conservation managers.

“Changes in how animals behave can scale up to affect wider ecosystem dynamics,” said Dr Keith. “But there are simply not enough humans to gather the behavioural data to see this happening in real time, so animal behaviour remains poorly integrated into assessments of ecosystem health.

“Through this project we will enhance monitoring capability through AI-driven approaches, alongside improved understanding of the underlying ecological mechanisms, providing new and faster ways to act.”

By integrating video, sound and environmental data, the projectwill deliver new insights into how environmental disturbance reshapes behaviour and how those changes can disrupt ecosystem function. This understanding will allow ecological risk to be anticipated and managed proactively.

Coral reefs are behaviourally rich, highly diverse and globally threatened ecosystems where behaviour is tightly linked to ecosystem function. Crucially, their clear water and abundant life also provide an ideal environment to develop computer vision models.

These insights and tools will then be adapted and evaluated in the more challenging conditions of UK coastal waters and marine protected areas, strengthening national capacity for marine biodiversity monitoring and management, and enhancing ecosystem resilience.

BEACON brings together an international network of partners. Prof Sergio Floeter (Brazil) and Dr José Paula (Portugal) contribute reef fish expertise and access to video archives spanning multiple biogeographic regions. The Centre for Environment, Fisheries and Aquaculture Science, Natural England, and the Seychelles Parks and Gardens Authority are co-designing the tools from the outset to ensure they meet the needs of conservation practice and UK marine policy. Together, these partnerships will ensure that what BEACON develops is deployable, useable, and as cheap and accessible as possible in the real world.

Dr Keith emphasised that "from the outset, BEACON will be transparent about what our tools can and cannot do, the costs and limitations of using AI, and who benefits from these decisions. It is crucial that we develop environmentally and socially responsible tools that conservation practitioners can trust and use with confidence.”

Professor Charlotte Deane, Senior Responsible Owner for the UKRI IS8 AI Programme and Executive Chair of EPSRC, said:

“AI has enormous potential to transform the way we understand and tackle some of the biggest challenges facing society. These fellowships will give outstanding researchers the freedom and support to push the boundaries of AI itself, while also applying it in new ways across a wide range of disciplines.

"They are also part of UKRI's long-term commitment to AI, building on sustained investment that is helping to strengthen the UK's position as a global leader in AI research and innovation.

“By investing in exceptional people and ideas, UKRI is helping to build the expertise and capability the UK needs to remain at the forefront of AI research and innovation, and to turn advances in AI to advance knowledge, improve lives and drive growth across the UK.”

Back to News