Can Past Mass Extinctions Predict Today’s Species Extinction Risks?

A new global analysis of tropical forest mammals is challenging one of conservation science’s most common assumptions: that the traits making species vulnerable to extinction remain consistent across time and place. By combining fossil evidence, species-distribution data and thousands of wildlife-camera images, an international research team has found that the characteristics associated with extinction risk over the past 130,000 years do not always predict which mammals are most vulnerable today. The finding could reshape how scientists identify species in urgent need of protection, particularly among the many animals whose populations remain too poorly documented for conventional assessments.

The study, published in Science Advances, examined 210 mammal species recorded across 64 tropical forest sites in Africa, Asia and the Americas. The dataset included animals ranging from porcupines, anteaters and monkeys to elephants, bears and large carnivores. Of the species analyzed, 199 still survive, although some have disappeared from portions of their former ranges. Eleven species are extinct, including giant pangolins that reached lengths of approximately nine feet, saber-toothed cats weighing as much as half a ton, and mastodons. By comparing these lost species with surviving mammals, the researchers reconstructed broad patterns of vulnerability before and after the global expansion of modern humans.

The urgency of the research is underscored by the current state of global biodiversity. Of roughly 6,000 known mammal species, approximately one in four is considered on a path toward extinction within the next century, according to the International Union for Conservation of Nature. At the same time, nearly 770 mammal species evaluated by the IUCN are classified as “data deficient.” This designation does not mean that the animals are safe; it means that scientists lack sufficient information about their abundance, geographic distribution or population trends to determine their level of risk. For these species, historical and comparative evidence can provide an important provisional guide.

Over the long term, the analysis identified a recognizable extinction-risk profile. Mammals with larger bodies, smaller brains, carnivorous diets and slow reproductive rates were generally more likely to disappear. These traits can create several biological disadvantages. Large-bodied animals often require more food and larger territories, making them sensitive to habitat loss and resource shortages. Carnivores may occur at naturally low densities because energy declines at each step of the food chain. Species that reproduce slowly cannot rapidly replace individuals lost to hunting, disease or environmental disruption. The researchers found that these associations appeared at local, regional and global scales, suggesting that they reflected broad ecological processes rather than isolated regional effects.

The role of brain size added a more complex dimension. Across the 130,000-year record, relatively large-brained mammals appeared to have an advantage over smaller-brained species. Larger brains may support behavioral flexibility, learning and the ability to respond to changing environments. Yet some research on present-day mammals has reported the opposite pattern, suggesting that large brains can become a liability under modern conditions. The energetic cost of maintaining a large brain is substantial, and species with advanced cognitive abilities may also mature later or produce fewer offspring. These competing pressures illustrate why a trait that appears beneficial over evolutionary time may not provide protection against the specific threats created by modern human activity.

The contrast became especially clear when the researchers compared the long-term fossil patterns with recent evidence gathered from automatic camera traps and other contemporary sources. Traits associated with vulnerability in the past did not consistently identify the mammals most likely to persist today. Large body size and slow reproduction, for example, were linked with resilience in the modern camera-trap dataset rather than with elevated risk. This result does not mean that elephants, large primates or other slow-breeding mammals are universally secure. Instead, it indicates that present-day survival patterns can be strongly shaped by where animals are observed and which human pressures they face.

One possible explanation is that many of the camera traps were placed in protected areas, where hunting and habitat destruction are less intense than in unprotected landscapes. In such locations, large mammals may benefit from reduced exposure to direct persecution and from conservation programs designed specifically to protect them. Their size and long generation times could then appear to be associated with survival, even though those same characteristics may make populations difficult to recover after severe declines. The researchers caution that this represents a potential sampling and context effect: a species may appear resilient inside a protected reserve while remaining highly vulnerable across its wider range.

The study’s central message is therefore not that historical extinction predictors have failed, but that they must be interpreted within a temporal and spatial framework. The ecological forces that drove extinctions thousands of years ago were not identical to those operating today. Ancient pressures included climate shifts, natural environmental changes and human expansion into previously isolated regions. Modern mammals face those pressures alongside road construction, agricultural conversion, illegal hunting, climate change, emerging diseases and the fragmentation of habitats. These threats can alter the importance of biological traits, sometimes reversing relationships that appear stable in the fossil record.

For conservation organizations, the findings offer both a warning and a practical opportunity. Historical patterns can help scientists prioritize poorly understood species, but they should not replace current population surveys, habitat monitoring and local ecological data. The most effective assessments will combine evolutionary history with information about contemporary threats, protected-area conditions and regional differences in human activity. “We can’t just assume that large-scale patterns apply at smaller scales,” said Lydia Beaudrot of Michigan State University, one of the study’s lead authors. “We have to take the temporal dimension into account.” As extinction accelerates worldwide, recognizing when old rules no longer apply may be essential for deciding which mammals still have a realistic chance to survive.

News Publication Date: 15-Jul-2026

Web References: https://www.science.org/doi/10.1126/sciadv.aeb7543; https://www.iucnredlist.org/statistics; https://www.iucnredlist.org/; https://www.wildlifeinsights.org/team-network

References: Schowanek, Simon D., Douglas Sheil, Lydia Beaudrot, Pierre Dupont, Santiago Espinosa, Vittoria Estienne, Julia E. Fa, Jonas Geldmann, Patrick A. Jansen, Steig E. Johnson, Francesco Rovero, Fernanda Santos, Asuncion Semper-Pascual, Andrea Fernanda Vallejo Vargas, Jorge A. Ahumada, Emmanuel Akampurira, Rajan Amin, Robert Bitariho, Adeline Fayolle, Davy Fonteyn, Ilaria Greco, Marcela Guimaraes Moreira Lima, Matthew Scott Luskin, David Kenfack, Emanuel H. Martin, Eustrate Uzabaho, Cedric Vermeulen and Richard Bischof. “Predictors of extinction risk in large tropical forest mammals: from global to local.” Science Advances, 15 July 2026. DOI: 10.1126/sciadv.aeb7543.