A new study highlights a growing risk for plant species globally. According to the study, more than 15,000 plant species have a high probability of being considered threatened or near-threatened under a new model used to predict conservation status. The model details the predicted levels of risk to plants worldwide and was published as part of a study to help governments and resource managers evaluate where conservation resources are needed most.

The model was built by a research team from the University of Idaho, University of Maryland, Radford University and Ohio State University and published recently in the Proceedings of the National Academy of Sciences.

The International Union for Conservation of Nature’s Red List of Threatened Species is a tool for researchers and policymakers working to limit the species in danger of being lost around the globe. A new approached developed by the University of Idaho and Ohio State uses machine learning and open-access data to predict plant species that could be eligible for at-risk status on the IUCN Red List.

Adding a single species to that list requires hours of expensive, rigorous and highly specialized research. Many known species have not been formally assessed by IUCN and ranked from “least concern” to “critically endangered.” Only about 5% of currently known plant species appear on the IUCN Red List in any capacity.

Letting the machine do the work
To build the model, the research team created and trained a machine learning algorithm to assess more than 150,000 species of plants from all corners of the world, making this one of the largest assessments of conservation risk to date. The researchers trained the model using open-access data from the relatively small group of plant species already on the list, and then applied the model to thousands of plant species that remain unlisted.

Anahi Espindola, who worked on the project as a doctoral researcher and is now an assistant professor at the University of Maryland, explained that this approach was not meant to replace formal assessments using IUCN protocols. “It’s a tool that can help prioritize the process,” Espindola says. “Ultimately, we hope it will help governments and resource managers decide where to devote their limited resources for conservation. This could be especially useful in regions that are understudied.”

The model predicted about 10% of total plants assessed by the team have a high probability of qualifying as “near-threatened” or worse. Maps of the data indicate at-risk species tend to cluster in areas already known for their high native biodiversity, including Central American rainforests and southwestern Australia. The model also flagged regions including California and the Southeast U.S., which are home to a range of endemic species not naturally occurring anywhere else on Earth.

“Although our primary goal was to help prioritize the process for ranking species, identifying geographic areas with high concentrations of potentially at-risk species was an added bonus,” says David Tank, associate professor in the University of Idaho’s Department of Biological Sciences.

The model also found a few surprising areas not known for their biodiversity, including the southern coast of the Arabian Peninsula, with a high number of at-risk species. Espindola hopes the study method can help to fill in some of these knowledge gaps by pointing out regions and species in need of further study.

The study was funded under a National Science Foundation grant.

Source: University of Idaho