In an urgent effort to protect freshwater fish from extinction, researchers have dedicated five years to developing a cutting-edge AI-based model. This innovative approach focuses on identifying potential threats to fish species before they reach endangered status.
“Often, people step in to save species when it’s already too late,” explained Ivan Arismendi, an associate professor in the Department of Fisheries, Wildlife, and Conservation Sciences at Oregon State University. “Our model allows decision-makers to allocate resources proactively, preventing species from becoming endangered in the first place.”
The findings of this research were recently published in the journal Nature Communications.
Alarmingly, nearly one-third of freshwater fish species are at risk of extinction, presenting threats not only to food supplies and ecosystems but also to recreational activities. The new model utilizes a machine learning framework to identify potential dangers facing over 10,000 freshwater species globally. Most of the species included in this model may still be preserved before they reach a critical state.
By analyzing 52 different variables that go beyond traditional assessments—such as damming, habitat degradation, pollution, economics, and invasive species—the model can more effectively identify threats. This innovative tool relies on publicly available data, making the process of safeguarding freshwater fish more efficient and cost-effective.
“This approach uses new metrics to determine what strategies are effective in preventing species from becoming listed as endangered,” noted Christina Murphy, a U.S. Geological Survey assistant unit leader for the Maine Cooperative Fish and Wildlife Research Unit and an assistant professor at the University of Maine. “This means that wildlife managers may be able to protect a significant number of fish populations.”
The tool promotes proactive conservation efforts by recognizing ecological, environmental, and socioeconomic patterns that support fish populations, enabling conservationists to implement targeted measures that benefit multiple species concurrently.
“The key insights are the socioeconomic factors influencing conservation effectiveness; we are much better at identifying what helps species thrive rather than what threatens them,” Murphy emphasized. “Using historical successes, managers can establish new conservation programs since many species share effective strategies.”
Researchers gathered data from 12 publicly accessible sources, predominantly from the International Union for Conservation of Nature.
They built and fine-tuned an AI system capable of analyzing millions of nonlinear interactions among species, effectively pinpointing those at immediate risk and the underlying causes of these threats. This platform allows users to investigate the factors contributing to vulnerability and to assess whether similar risks might impact other species that are not yet in danger. Furthermore, the research team validated their model against existing conservation assessments.
The team believes that their tool can serve as a valuable asset for conservation and regional planning initiatives, with aspirations to extend its application to protect birds, trees, and other flora and fauna.
“Our findings illustrate that conservation operates similarly to human health: the indicators of ‘well-being’ are often more reliable than the various paths leading to decline,” said J. Andres Olivos, a post-doctoral scholar at Oregon State. “For freshwater fishes, safe environments tend to be predictable, while the risk of extinction stems from an array of complex threats.”
The project was initiated by Murphy in 2020 while she was a postdoctoral researcher at Oregon State, collaborating with Arismendi and Olivos alongside scientists from the USGS, the U.S. Forest Service, and the University of Girona in Catalonia, Spain.
For more information, refer to the University of Maine press release highlighting this research.
