Ships are making too much noise in the Strait of Gibraltar, which could be causing a critically endangered population of long-finned pilot whales to have trouble hearing each other. To better understand the whales’ struggle, scientists recorded their calls as well as the background noise. The team discovered that the whales can increase the volume of some calls, but they are already shouting as loud as possible to find each other after diving deep to search for food.

A new study finds that large language models (LLMs), used with straightforward prompting, perform poorly on routine number-crunching tasks that hospital administrators depend on every day to track patients and allocate resources. The findings were published this week in the open-access journal PLOS Digital Health by Eyal Klang of the Icahn School of Medicine at Mount Sinai, New York, USA, and colleagues.

In a process analogous to how solids melt into liquids, the electrons in many different metals form crystal-like patterns that can deform and melt, opening new pathways for neuromorphic computing and superconductors, University of Michigan Engineering researchers have found.

Plants underpin the majority of life on Earth, yet climate change is rapidly reshaping their habitats and elevating their extinction risk in largely unknown ways. Now, in two studies, researchers use large-scale evolutionary modeling and climate projections to fill these gaps, revealing substantial losses of plant diversity and identifying priority species and limits of conservation strategies. Understanding the endangered statuses of plants is crucial to understanding the biosphere’s future and guiding effective conservation efforts. Yet plants are largely absent from global biodiversity assessments. Although more than two in five plant species are estimated to be threatened with extinction, only about 20% have global International Union for Conservation of Nature (ICUN) Red List assessments. Across two studies, Félix Forest and colleagues and Junna Wang and colleagues present two distinct predictive approaches that address the current gaps in available plant biodiversity data.

 

Using data from the latest version of the Evolutionarily Distinct and Globally Endangered index (EDGE2) and computational modeling, Forest et al. reconstructed large-scale evolutionary trees encompassing all 335,497 known angiosperm species and combined them with projected risk data. This allowed the authors to identify species that are both evolutionarily unique and at risk of extinction. According to the findings, roughly 21% of angiosperm evolutionary history is at risk of extinction. Moreover, the study identified 9,945 angiosperm species that, if conserved, would most effectively preserve the deep evolutionary heritage of plant life.

 

In another study, Wang et al. analyzed the geographic distributions of 67,664 vascular plant species to forecast how climate change may alter their habitats over time. By comparing the pace of environmental change with each species’ ability to relocate, the authors assessed whether plants could successfully track shifting conditions or face increasing extinction risk. Wang et al. found that the primary driver of plant extinction is not a plant’s limited ability to shift its ranges, but rather the widespread loss of suitable habitats caused by climate change. Using several greenhouse gas emission scenarios for the end of the century, the findings predict that between 7% and 16% of the species examined could face a high risk of extinction, as most of their viable habitats disappear. Although shifts in species’ geographic ranges are unlikely to substantially reduce global plant extinctions, they are expected to increase local plant diversity across roughly 28% of the Earth’s land surface. According to Wang et al., facilitating range shifts via conservation efforts may help sustain or even enhance regional richness. However, it does little to prevent the broader, worldwide loss of plant species.

 

“Although Forest et al. and Wang et al. used different scales of time and space and studied different, but largely overlapping, groups of plants, both studies revealed that plant extinctions do not occur randomly across geographical areas,” write Rosa Scherson and Federico Luebert in a related Perspective. “Large-scale predictive models such as those developed by [the authors] are a valuable tool to enable timely actions that cannot wait until complete knowledge about biodiversity loss is achieved.”